Python Algorithmic Trading Course – Massive, SnapTrade & Alpaca Integrations

[00:00] In this practical [00:01] you are on course [00:01] will use [00:02] Python for creation [00:04] full-fledged [00:04] algorithmic [00:05] trading systems [00:06] paper [00:07] tools from scratch. [00:09] You will configure [00:10] uninterrupted [00:11] connecting conveyor [00:12] market data from Massive [00:14] for safe [00:14] portfolio execution [00:16] via SnapTrade. Main [00:18] program logic [00:19] calculates 12-month [00:21] momentum indicators [00:22] for a universe of 50 [00:23] shares for generation [00:25] automated [00:26] trading signals. These [00:28] signals then [00:29] automatically [00:29] are directed to [00:30] simulated [00:31] brokerage account [00:32] Alpaca for [00:33] risk-free [00:34] implementation. This course [00:36] became possible [00:37] thanks to a grant from [00:38] SnapTrade and Massive. Nick [00:40] McCollum created [00:41] this course with the help of [00:42] Katie Charbonneau. [00:43] Hi, I'm Katie from SnapTrade. [00:45] I will show you how. [00:46] customize yours [00:47] SnapTrade account [00:48] . So, the first thing you need to do is [00:49] need to do this [00:50] go to the website [00:51] snaptrade.com and click [00:52] "Start" button. Now [00:53] you can choose " [00:54] commercial" and [00:55] press " [00:56] Continue". Now you [00:57] can you fill this out [00:58] information and [00:58] press the " button [00:59] Register." [01:00] So, now I have created [01:01] your account, [01:02] and the next thing I know [01:02] need to do this [01:03] confirm your [01:03] email address [01:04] mail. So I'm going [01:05] to use yours [01:06] by phone. I will press [01:07] by email [01:08] with the inscription " [01:08] Confirm your [01:09] email address [01:10] Shapeshift" and I'll go for [01:11] link in it. [01:12] You will be redirected to [01:13] page from [01:13] message that [01:14] your email [01:15] address successfully [01:15] confirmed. As [01:16] only you will see it [01:17] , you can update [01:18] page and see, [01:20] that your email [01:21] Address confirmed. [01:22] So, I'll click " [01:23] API keys", and before I [01:24] I can set this up. [01:26] we need [01:26] to adjust [01:27] two-factor [01:28] authentication. So I [01:29] I will press " [01:30] Security settings [01:31] ". I will use it. [01:33] with your Google app [01:35] Authenticator. So I will open [01:37] and scan the QR- [01:38] code. And from there I can [01:41] enter the code. Therefore, [01:43] now you can [01:44] to see that mine [01:45] Authenticator app [01:46] is on, so this is [01:47] configured. So, I [01:48] I can go back to [01:49] API keys. Therefore, [01:50] from here I will click " [01:51] Generate key. [01:52] Now we are here and you [01:53] you can copy [01:54] your ID [01:55] consumer and [01:56] save it. [01:56] Copy yours [01:57] identifier [01:58] consumer and [01:58] save it. Very [01:59] it is important to keep this [02:00] in a safe place, [02:01] because you are not [02:02] you can see this [02:03] again after closing. [02:04] So, once you do this [02:05] save, you can [02:06] press "I" [02:07] copied yours [02:08] identifier [02:09] consumer". That's all. [02:11] So, I'm handing you over. [02:12] Nick, who will conduct [02:13] you through the rest of this [02:15] course. [02:16] Okay, so in this one [02:16] part of the video I [02:17] I will show you how. [02:18] create an account [02:19] Massive recording so you [02:20] could generate [02:21] the Massive API key that we [02:22] then [02:22] we will use [02:24] later in the video for [02:25] data acquisition [02:25] stock market with [02:26] Massive API. The first thing you [02:28] need to do this [02:29] go to massive.com, and [02:29] on their main [02:30] page at the top [02:31] right corner you [02:32] notice that they have [02:33] this blue button " [02:34] create an account [02:35] recording". So you [02:36] will need [02:36] press it now. [02:37] After you [02:38] register, there is [02:39] several different [02:39] registration methods. [02:40] You can [02:41] register [02:41] via GitHub via Google [02:43] or by using [02:43] email and [02:44] password. We will [02:45] use [02:46] email. I [02:50] I think that this [02:50] sends an email [02:51] letter from [02:52] confirmation on [02:52] my address [02:53] email. [02:54] That's right. So, let's [02:55] let's find it now. [02:57] It seems that the link [02:58] for confirmation [02:59] opened in a new [03:00] browser, so I [03:01] I'll drag it here. [03:02] Okay, and here we are [03:04] here. So, Massive is quite [03:05] simple in [03:06] use. Here [03:07] many settings, [03:08] but in reality all that [03:09] we need this [03:09] opportunity to get [03:10] the key is right here. So we [03:12] click on this. [03:13] It looks like we already have it. [03:14] API key for [03:14] default. You [03:15] you want to take it [03:16] and use in [03:18] in your project in [03:18] future. So I [03:19] I'm going to return. [03:20] this is so that no one who [03:21] watch the video, no [03:22] could it [03:22] use. But, [03:24] yes, here's what you need [03:24] need to be done to [03:26] get your API key [03:27] Massive. And make sure [03:28] that you saved it for [03:29] then, because we [03:30] we will use [03:31] this is in the project that we [03:32] we create. Good, [03:33] Next we need [03:35] create an account [03:36] recording in Alpaca. So, Alpaca [03:38] —this is, uh, stock [03:39] brokerage company. [03:41] They have a public [03:42] stock brokerage [03:43] company that can [03:44] register [03:45] everyone, but what is important [03:46] for this video, they have [03:47] there are paper trading [03:48] bills. So, you [03:49] you will be able to follow [03:50] according to this educational [03:51] manual, in fact [03:52] without replenishing [03:53] brokerage account [03:54] and not carrying out [03:55] no bidding, [03:56] using [03:56] real money. Therefore, [03:58] the first thing you need to do [03:59] need to do this [04:00] go to alpaca.markets. [04:02] This is not alpaca.com. Believe me, [04:04] I was there. It [04:05] real website, [04:06] related to [04:07] alpacas. So, when [04:08] you are on [04:09] alpaca.markets, at the top [04:10] right corner you [04:11] you will see a button [04:12] registration. They have [04:13] two different APIs: [04:14] trading API and [04:15] Broker API. You [04:16] will need [04:17] register in [04:18] trading API. As [04:19] only you here [04:20] you will find yourself, system [04:21] just collect some [04:22] information about you. [04:23] So, my name is, oh, [04:35] need a stronger one [04:36] password. Steeply. How are you? [04:40] Did you hear about us? [04:41] Let's say YouTube, because, [04:42] probably many of [04:43] you are watching this on [04:44] YouTube. So, we [04:45] click " [04:46] register". Oh, [04:47] I need [04:47] choose a country from [04:48] this dropout [04:49] list, I think. [04:52] Interesting. Maybe they [04:53] do not support [04:54] Canada, so I'm a little [04:55] I'll cheat and say that I [04:57] from the USA. Register [04:58] . Okay, I got it. [04:59] confirmation [05:00] email, like [05:02] and before, so let's [05:03] Let's find it. That's it. [05:04] It says: "Therefore," [05:05] it opened in [05:06] another window, but I [05:07] I'll move it here, [05:08] so that you can do this [05:09] see. So, there [05:10] It is written that your [05:11] email address [05:11] confirmed. Now [05:13] you can enter. We [05:14] let's move on [05:14] login page and [05:15] Let's go in. And then I [05:21] I'll press this little [05:22] "remember" field [05:23] me". Okay, let's go. [05:25] let's continue. Good, [05:26] so we are on the main [05:27] on the Alpaca page. Here [05:30] immediately worth it [05:31] pay attention to this [05:32] blue message. You [05:34] you will see what is there [05:34] It is written: "You [05:35] do you trade paper [05:36] prices, and real [05:37] money is not [05:37] are used." [05:39] This was important for [05:39] us during creation [05:41] this video because [05:42] we wanted everyone [05:43] could participate, [05:44] if he wants, [05:44] regardless of your [05:45] financial condition. Em [05:46] if you want [05:47] go to real [05:48] trade, you can [05:49] press this [05:49] drop-down menu in [05:50] upper left [05:51] corner and go from [05:51] paper trade [05:52] on an individual basis [05:53] trade. You can [05:54] to see that this is fresh [05:55] paper trading [05:56] the account has [05:57] false balance in [05:58] amount of 100,000 dollars. [05:59] So, that's where we are [06:01] we will base our [06:02] project. Hmm, there are others. [06:03] interesting things in [06:04] interface. You can [06:06] see your positions, [06:07] your orders, your [06:08] activity, own [06:10] balance sheets and something else. You [06:11] you can see that everything [06:12] it is closed because [06:13] the system wants me to [06:13] configured the application [06:14] for authentication. [06:15] So I'll do it now. [06:17] Actually, I'll do it. [06:18] off-screen so as not to [06:19] to worry about [06:19] hiding their [06:20] credentials or [06:21] something like that. But [06:22] You guys seem to [06:23] will need [06:24] customize yours [06:24] authenticator to [06:25] really take [06:26] participation in this video. [06:27] If you have never before [06:28] they didn't do this, [06:29] there are many different [06:30] programs that you [06:31] can you [06:31] use for [06:33] this, such as Microsoft [06:34] Authenticator, Google Authenticator, [06:35] Authy, Bitwarden. Majority [06:36] programs for [06:37] password management, [06:37] what kind of people [06:38] use [06:39] today, have this [06:39] built-in function like [06:40] -from 1Password or LastPass. [06:41] So, make sure [06:42] what did you set up [06:43] your authenticator, [06:44] before moving on [06:44] to the next step. [06:45] Otherwise we [06:46] we can delve into [06:47] project. Okay, for [06:48] outside the screen I [06:49] set up your [06:50] addition- [06:51] authenticator and [06:52] I just would like to [06:53] to briefly introduce [06:54] you with the Alpaca interface [06:55] . We will be here long enough. [06:57] long, until [06:57] we are developing our [06:58] project because we [06:59] we are going to create [07:00] application that [07:01] will allow us [07:02] place deals. [07:03] Then we will move on to [07:04] Alpaca interface and [07:05] let's look here to [07:06] to make sure that we [07:07] we place the necessary [07:09] agreements. So, this is [07:10] home page. You [07:11] get balance [07:12] your account, and [07:12] then yours [07:13] historical balance. You [07:14] you can see that [07:15] because it is new [07:16] paper trading [07:17] account, and we never [07:18] not really [07:18] placed any [07:19] agreements or something like that [07:20] , the balance is constant with [07:21] sometimes at the level of 100,000 [07:22] dollars. Other main [07:23] places we will be [07:24] to search, are located [07:25] in the sidebar here. [07:27] So, first this [07:28] was closed. I [07:29] opened submenu [07:30] account. AND [07:31] then you can [07:32] to see yours [07:33] positions. Therefore, [07:34] obviously we don't have [07:35] positions. You can [07:36] see your orders. [07:37] Obviously, we don't have [07:38] warrants. Activity, [07:39] balances, and then [07:40] some options [07:41] configurations. So, in [07:42] we are on [07:43] fractional trade and [07:44] short positions. AND [07:45] then we have [07:45] pricing and [07:46] trade support [07:47] third party options [07:48] equal. So, here [07:49] a lot of useful things. We [07:51] we will see more about [07:52] this is when the project will be [07:53] built during [07:54] the rest of this video. We [07:56] we are going to start [07:57] creating our [07:58] first project. AND [07:59] before we [08:00] let's delve deeper and [08:01] let's start writing code, I [08:02] I would just like to [08:03] to walk and give you [08:04] a kind of idea [08:06] about how [08:07] the project will look like [08:08] after him [08:09] completion. So, this is [08:10] information panel. [08:13] And the fact that we are here [08:13] we create, -this [08:14] trading system [08:16] investing in [08:17] momentum, which [08:17] allows you to take [08:19] a whole range of shares, [08:20] rank them by 12- [08:21] lunar momentum, [08:23] that is, in essence, for [08:24] by changing their price for [08:25] the last 12 months. You [08:27] rank all stocks in [08:28] the universe behind this. AND [08:29] there are many [08:30] academic sources [08:31] information about the idea, [08:33] what if you buy [08:34] stocks with the highest [08:35] momentum and [08:35] potentially selling [08:37] stocks with the lowest [08:38] momentum, you [08:39] will receive [08:39] extremely high [08:41] profit. This [08:41] the strategy we [08:43] we develop, [08:47] inspired by the book " [08:48] Quantitative Impulse" [08:50] Jack Vogel and Wes [08:51] Gray. They manage [08:53] by a quantitative firm [08:54] investing in the USA, [08:55] which implements [08:56] many of these [08:57] quantitative strategies [08:58] , but it's kind of [08:59] inspiration for this. [09:00] So, I'll show you all [09:01] different things that can [09:02] to do this little one [09:04] software application, and [09:05] then we step by step [09:06] Let's consider step by step. [09:07] him and we will create everything [09:08] it. So, this app [09:09] in the main [09:10] divided into four [09:11] sections. Do you have [09:12] toolbar, [09:13] Do you have indicators? [09:14] momentum. So, here [09:15] no data yet. [09:16] We will consider this for [09:18] a second. And then there is [09:19] trading signals on [09:20] based on these [09:20] momentum indicators [09:21] . Finally, there is a section [09:23] for management [09:24] your portfolios. [09:25] So, this is the part, [09:26] where will we be [09:27] use [09:28] SnapTrade to actually [09:28] connect this [09:29] addition to your [09:30] brokerage account [09:31] for the implementation [09:32] bidding and [09:32] monitoring your [09:33] balance sheets. So, yes. [09:35] First of all, let's [09:36] let's enter some data in [09:37] this thing. You [09:38] will need [09:39] press this button in [09:39] upper right [09:40] forged, to count [09:41] momentum. Uh, and this [09:42] just a hint: "You [09:43] want to list [09:44] momentum estimates for [09:44] all shares? This can [09:45] to take a few [09:46] minutes." Let's [09:47] Let's do it. And then we [09:48] we get a hint, [09:49] that momentum estimates [09:51] successfully listed [09:52] for 50 shares. You [09:53] you'll see when we do it [09:54] actually create, [09:55] but we are tough [09:56] encoded a small [09:57] a universe of 50 stocks, [09:58] which consists of [09:59] Investly universe for [10:00] this application. Um, [10:01] okay, so here's what we [10:02] we are considering. You see, [10:04] we have two here [10:05] main tables: 10 [10:06] best promotions [10:07] momentum and 10 [10:08] worst stocks [10:09] momentum. Um, 10 [10:10] the best ones are a bunch... [10:11] So, uh, it's March. [10:13] 2026. Top 10- [10:14] this is, in fact, a pile [10:15] artificial intelligence shares [10:16] intelligence and [10:16] technological [10:17] companies, because [10:18] lately we [10:18] we are experiencing a great [10:19] artificial boom [10:20] intelligence. So, Google [10:22] and AMD, Intel, Nvidia, Tesla, etc. [10:24] . I'm a little surprised, [10:26] I see promotions here. [10:27] blue chips, such [10:29] like Johnson & Johnson. But this one [10:30] impulse is, in essence, [10:31] their 12-month price list [10:32] yield. So, this is [10:34] decimal. You [10:35] you can also [10:35] express it in [10:36] percentage. This means [10:38] that Intel grew by 98.5% over [10:39] last year, etc. AND [10:41] then, similarly, in [10:42] We have 10 shares with [10:43] the lowest indicator [10:43] impulse. So, United [10:45] Health fell by 45% [10:46] the last 12 months. Adobe [10:48] fell significantly. Uh, [10:49] a lot of these things are very [10:50] fell, I think it's possible [10:52] to say, because of fears [10:53] before artificial [10:54] intelligence, etc. Therefore [10:55] , SAS, etc. Uh, Salesforce, [10:57] Adobe is good [11:00] examples of this. AND [11:01] then you can see [11:02] that we don't have [11:02] active portfolios and [11:03] no events [11:04] rebalancing. AND [11:05] then the only thing I don't [11:06] I remembered, this is what [11:07] there is some on top [11:07] information about, uh, [11:09] average momentum and [11:10] median impulse, e- [11:11] e, and also about our [11:12] briefcase, etc. Therefore, [11:13] if you want [11:14] to see, uh, [11:15] impulse indicators [11:15] for the entire portfolio [11:16] or rather, for the whole [11:17] systems, you can [11:18] go to tab [11:19] impulse indicators, [11:20] scroll down, you [11:21] you can see them [11:22] everyone is in the ranking. Top 10 [11:23] –all green, bottom 10 [11:24] –red, and [11:25] theoretically one of [11:26] strategies that you [11:28] could be applied in [11:29] this universe, is [11:30] open a long [11:31] top 10 position and [11:33] open a short [11:34] position in the bottom 10. [11:35] Hmm, yes. There is also [11:36] conversion button, [11:37] if you come back [11:38] tomorrow and you need [11:39] fresh data, you can [11:40] to use it. AND [11:41] the last one is the page [11:42] trading signals. [11:43] So, this is where you are [11:44] you can get [11:45] trading recommendations [11:46] regarding all shares in [11:47] the universe. But first [11:48] than we can do this [11:49] to do, we need [11:50] connect the portfolio. [11:51] You can see, [11:51] if we try [11:52] do it now, uh, [11:53] it will say: "You [11:53] want to generate [11:54] new trading signals [11:54] based on current [11:55] momentum data? [11:56] This may take [11:57] a few minutes. You [11:57] you will get an error, [11:58] because we don't have [11:59] active portfolios. [12:00] Let's fix this, [12:01] by connecting [12:02] account [12:03] via SnapTrade. Therefore, [12:04] press the button « [12:05] Portfolios" on the left. [12:06] Click "Create" [12:07] your first briefcase [12:08] ". This will guide us. [12:09] through the OAuth process to [12:10] connect this [12:11] program to our [12:12] paper [12:12] account [12:13] Alpaca with the help of [12:14] SnapTrade. So, we [12:15] let's call this portfolio [12:17] "Momentum Portfolio" [12:19] 2026". And ours [12:20] investment [12:21] the strategy is to [12:22] in order to buy 10 [12:23] best promotions [12:24] momentum. Good. [12:28] So, this is integration. [12:29] live trading. There is here [12:30] a few warnings. THERE ARE [12:32] several supported [12:33] brokerage offices. [12:34] It is now limited. [12:35] only Alpaca for this [12:36] manual. And then we [12:38] click "Secure" [12:39] connection via [12:40] SnapTrade. Click here. [12:42] It appeared on [12:42] to my friend [12:43] monitor, but we [12:44] we get a pop-up [12:44] a window that looks [12:45] like this. Allow [12:46] I have to do it. [12:47] a little bigger. Now [12:48] I already...I entered the [12:49] your account [12:50] Alpaca recording on this [12:51] computer, therefore, [12:52] probably, I have [12:52] saved token [12:53] device in Alpaca, and this [12:54] means that he, [12:55] probably not [12:56] will ask me [12:56] enter your name here [12:57] user or [12:58] password, but usually [12:59] it is being done. So, this is [13:00] still Snap Trade, he [13:01] just telling [13:02] to you, what is here? [13:03] is happening. We [13:04] we are going to finish [13:05] process [13:06] authentication by [13:07] using Alpaca paper. [13:08] And now we are on the web- [13:09] on the Alpaca website. So, I [13:10] I think if you [13:11] did this because [13:11] haven't been in for a long time [13:12] system, he first [13:13] will ask you to enter [13:14] your username [13:15] and password, then [13:15] redirect you to [13:16] this page. Therefore, [13:17] which paper [13:18] Alpaca account [13:18] Do you want to connect? [13:19] Well, let's go. [13:20] Let's connect this one. [13:21] Paper accounting [13:22] record—the only one that [13:22] There are us. We press " [13:23] allow", and this [13:24] will return us first [13:25] to our app [13:26] after connection. [13:26] Boom. Connection [13:28] completed. So, this is [13:29] opened in a new [13:30] the window that I am in now [13:31] I will close. And then, if [13:32] I will refresh the page. [13:33] briefcases here, you [13:34] you can see our [13:35] momentum portfolio, [13:36] which we just did [13:37] connected. Perfectly. [13:38] Now we can [13:39] return to [13:39] shopping page [13:40] signals and actually [13:41] to make...some deals. [13:42] So, we press [13:43] this green button in [13:44] upper right [13:44] carved with the inscription " [13:45] generate signals". [13:46] The prompt says " [13:47] generate new ones [13:47] trading signals on [13:48] based on current [13:49] momentum data". [13:50] This may take [13:50] a few minutes." Yes. [13:52] So, we generated [13:53] there are 10 signals here. And this one [13:54] the system is configured [13:55] so that the data is in the mode [13:56] real time [13:57] may be a little [13:58] expensive, and they don't [13:58] included in [13:59] free plan, [14:00] which we have with Massive. [14:01] Therefore, I think, [14:01] usually you want [14:02] weigh the stock with [14:03] with the highest momentum [14:04] slightly higher than the stock [14:06] with the 10th largest [14:06] momentum, and make [14:08] weighted average [14:08] value. We did [14:10] it's a little easier. We [14:11] just getting together [14:12] invest 100 [14:13] dollars in each of 10 [14:15] best promotions [14:15] momentum. So, in [14:17] each column has [14:18] "buy" button. You [14:19] click "buy", [14:20] and it says:" [14:20] Okay, we're going. [14:21] buy Walmart. Size [14:22] deal is 100 [14:23] dollars ". Accurate [14:24] the number of shares will be [14:25] calculated during [14:26] execution from [14:27] using [14:27] current market [14:28] prices. And this is actually [14:30] takes place on [14:31] on the Alpaca side, not on the [14:32] us. So, we [14:33] we are going to him [14:34] perform. Boom. Therefore, [14:35] our purchase order [14:36] Walmart for $100 was [14:38] successfully submitted [14:38] to the broker. This is great. [14:40] Theoretically you could [14:41] to walk and perform [14:42] all these orders, but [14:43] instead of this [14:44] let's just [14:44] let's go to Alpaca and [14:45] we will make sure that this [14:46] the deal is posted [14:47] right. Good, [14:49] I need to enter [14:50] here is the password. So I [14:51] I'll just do it in [14:53] another window. I [14:53] I'll be back in a second. [14:58] Perfectly. So, I am now [15:00] entered Alpaca. You [15:01] you can see, uh, [15:02] that is, it's all paper [15:04] trade. Uh, if you [15:05] go to [15:05] account and [15:06] warrants, we must [15:07] see the warrant for [15:08] I'll buy Walmart for 100 [15:09] dollars. This is a warrant, [15:10] which we just did [15:10] placed. If we [15:12] Let's open it, you [15:13] you will be able to see [15:14] warrant. Now, in [15:15] Alpaca interface [15:16] some are missing [15:17] elements regarding [15:17] conditional agreements. So, you [15:19] you can see that in [15:20] it has no limit [15:21] prices, no stop price, [15:22] um, purchases, etc. And you [15:23] like: "Oh, where is 100 [15:24] dollars? "It's easy [15:25] to miss, it [15:26] right here. But this [15:27] the deal we just made [15:28] placed for 100 [15:29] dollars at Walmart. AND [15:31] in a similar way you [15:32] could easily buy [15:33] all the others from the top 10 and [15:34] create your own [15:35] own portfolio [15:36] impulse from zero, [15:36] using [15:37] software [15:38] the provision that we [15:39] we are going to create [15:40] in this project. Therefore, [15:41] here is your little one [15:42] preview [15:43] project. Stand up, [15:45] Take a walk, a little. [15:46] take a break, [15:47] make some coffee and then [15:48] come back. We [15:49] we are going [15:49] to delve deeper and [15:50] create this thing from [15:51] zero now. Good. [15:58] Now I will show you how [15:59] customize your [16:00] local environment [16:01] for starters, as well as [16:02] I will explain mine. [16:03] recording philosophy [16:04] this so that you can [16:05] to learn as much as possible [16:06] more while you watch [16:07] behind this. This [16:09] repository that [16:10] is behind [16:11] at nickmccollum/ [16:12] algorithmictrading, is a place, [16:13] where I thought a lot about [16:14] these projects, created [16:16] them in parts, [16:17] tested them and it can [16:18] to be final [16:19] source of truth about [16:21] of how it should be [16:21] look like a project. [16:25] In the future we [16:25] we will work on [16:26] thereby [16:27] repository, but with [16:28] added dash entry. [16:30] So, this is empty. [16:31] repository. I [16:32] I'm going to clone. [16:33] its in my local [16:34] development environment, [16:35] and we will collect it [16:36] in parts. I [16:37] I am going to record. [16:38] all the mistakes that I made [16:39] I do, and show [16:40] you how I fix them [16:41] . You will see the real [16:42] commit history in [16:43] repository you are using [16:44] you will be able to [16:45] to use to [16:46] see where I am [16:47] I decide to commit. [16:48] things that [16:48] commit message [16:49] create etc. And, by [16:51] essentially, this repository [16:52] will serve as a step-by-step [16:53] review for [16:54] actual recording, [16:55] which one are we going to [16:56] to do, uh, within the framework [16:57] this video. So, uh, [16:59] what do you want to do [17:00] locally, copy [17:01] this is, uh, sorry, [17:02] first create [17:03] your own [17:04] repository so that you [17:05] could do it [17:06] independently. [17:08] Copy this, uh, this [17:09] SSH address, and then [17:11] run get clone, and [17:12] then insert it. Uh, [17:15] you will have an empty [17:16] repository locally [17:17] which you can [17:18] open and [17:19] to start. Good, [17:20] So, I opened... [17:21] Cursor to new [17:22] repository that we [17:23] just created, which [17:24] is called [17:25] algorithmic [17:26] trade...Now we [17:30] need to be configured [17:32] our catalog [17:33] project. We need [17:34] customize requirements [17:36] Python and getting started [17:37] initial formation [17:38] project framework. Therefore [17:40] , the first thing we'll do, [17:41] this will create a new one [17:42] folder named " [17:43] Project one. "And then [17:45] "inside" the Project [17:47] one "we will create" [17:48] file named [17:49] requirements.txt. In this [17:52] files are stored [17:53] all your dependencies [17:55] Python. If you have switched [17:58] from the land of TypeScript or [18:00] Node.js, that's about it [18:02] same as package.json file [18:03] . When you [18:04] install [18:05] third-party dependencies, [18:06] you write them down here. [18:07] So, I open [18:09] your integrated [18:10] terminal. We [18:11] let's move on to [18:12] catalog" Project one [18:13] ". And then, the first thing we [18:16] we want to add to [18:17] of this requirements file, this is [18:18] Django. Django— [18:20] most popular web [18:21] Python framework. He [18:23] allows easy [18:24] process HTTP requests. [18:26] This allows [18:27] integrate with [18:28] databases, such as [18:29] like PostgreSQL, and it will be [18:30] the basis of what we [18:31] we use for [18:32] creating this [18:33] project. So, uh, we [18:34] we are going [18:35] install this in our [18:36] local ee [18:37] environment. Before [18:38] everything, uh, we [18:39] we are going [18:39] use [18:40] package manager under [18:41] called UV. Uh, so we [18:43] we are going to perform [18:46] UV pip install-r requirements .txt. [18:48] Now, I think it's us [18:50] asks: uh, he says, [18:52] file not found [18:54] requirements.txt. What do we have? [18:56] Hmm. Why doesn't he see? [18:58] ? Oh, it is located [19:02] out of catalog [19:03] project. Let's [19:04] let's move it [19:05] there. Good, [19:06] Let's move this. Good [19:07] , now we are going [19:09] run UV pip install -r [19:10] requirements.txt. This is beautiful. [19:11] request. Uh, let me. [19:12] I have to do it. [19:13] a little bigger. It [19:14] good request. He [19:15] says not found [19:16] virtual [19:17] environment, [19:17] run UV VNV to [19:18] create an environment. [19:19] So this is, uh, UV VNV, uh, [19:21] creation [19:21] virtual [19:22] environment simply [19:23] means that everyone [19:24] dependencies that we [19:25] we establish, will be [19:27] isolated from this [19:28] project. So, we [19:30] let's create a virtual [19:31] environment, then [19:32] we will establish [19:33] Python dependency. You [19:35] you can see that Django, [19:37] Asgi ref and SQL parse [19:38] installed. These two, [19:40] obviously, there must be [19:41] Django dependencies. [19:44] Now we can [19:45] create a framework for [19:46] our project. Django [19:48] comes with [19:49] built-in command [19:51] called start project, [19:52] which allows you to give [19:53] the project name, and [19:55] then Django will review [19:56] and will create all the different [19:57] files needed for [19:59] project launch. I [20:01] I'll show you how. [20:02] looks. Team [20:05] Django-admin start project, a [20:06] then you need [20:07] Give the project a name. [20:09] You saw from the description [20:10] project, what is this project [20:11] software [20:12] provision for [20:13] momentum trading- [20:14] rating, so we [20:15] let's call this project [20:16] momentum trader. I am right. [20:17] wrote? Momentum trader. Yes [20:19] . Okay, you'll see. [20:21] on the sidebar, boom, [20:22] a bunch of things just now [20:23] created. I actually [20:26] I will move the created [20:27] things on one level [20:29] above so that they are in [20:30] project one catalog. AND [20:32] now we can [20:33] start our server [20:35] Django locally and [20:36] to see how he is [20:37] looks. Excuse me, I [20:38] moved them to [20:39] start of the repository. [20:40] I want them to be [20:41] within the project [20:42] one. That's it. OK, [20:45] now we can [20:46] start our server [20:47] Django. So, the command for [20:50] this-Python manage.py uh run [20:52] server. Cool. Therefore, [20:57] a few things. On- [20:58] first, you can [20:59] see that the server [21:00] works, and every time [21:00] we make any [21:01] changes to our code [21:02] base, for example, if [21:03] we'll just add here [21:04] a little space, and [21:05] then we will save it, [21:06] you will see that [21:06] server immediately [21:07] will reboot. [21:08] Okay, so you can [21:09] to see that this is the first [21:10] once the server [21:11] started, and then [21:12] This was the second time. [21:13] when the server [21:13] started. This is the first. [21:16] Secondly, you can [21:17] see what Django [21:18] warns us that in [21:20] there are 18 of us [21:20] unused [21:22] migration files. It [21:24] means that migration [21:25] Django means change [21:26] database schemas, on [21:27] which yours works [21:29] Django project. So, if [21:30] you add a column, [21:31] this creates migration [21:32] Django. If you [21:33] delete the table, this [21:34] creates a Django migration. [21:35] Whenever you [21:36] make changes to the database [21:37] data, software [21:38] software [21:39] automatically creates [21:40] Django migration, and, [21:41] obviously, currently in [21:41] Django built-in 18 [21:42] migrations. To us [21:43] need to apply [21:44] them to our local [21:45] databases. Therefore, [21:46] the command for this is python [21:48] manage.py migrate. This is a list [21:51] of all the migrations that we [21:52] just applied to [21:53] our database. [21:55] So, here we go again [21:55] let's start our server, [21:57] and this warning [21:58] there will be no more. Therefore, [22:00] python manage.py runserver. [22:02] Okay, you see that [22:03] warning about [22:04] migration more [22:05] no. So, we can [22:08] visit our server [22:09] and see how he [22:10] looks. Let's [22:11] let's open it in [22:12] another window. I [22:13] I'll drag it here. [22:14] Perfectly. Therefore, [22:15] installation completed [22:16] successfully. Congratulations. It [22:19] there will be a web page, [22:20] which in the end [22:21] will turn into [22:22] Momentum Trading app, [22:23] which I was showing you [22:24] earlier. Okay, so, [22:27] now we will create [22:28] several Django applications. [22:31] There is a difference in the world of Django [22:33] between the project that we [22:34] just created, and [22:35] an application that is like [22:36] isolated part [22:38] project. So, if [22:40] think, for example, [22:42] if you create [22:43] DoorDash, you can have [22:45] one project under [22:46] called the DoorDash app, [22:47] and there is, sorry, a project [22:49] DoorDash, I have to say. AND [22:51] inside it [22:52] there can be two different [22:54] applications for [22:54] driver management [22:56] and buyers, [22:57] for example, in this [22:58] project we will create [22:59] two applications under [23:00] called trading and portfolio, [23:01] to accommodate different [23:02] things related to [23:03] project. Django, uh, has [23:05] built-in [23:06] functionality for [23:07] automatic [23:08] loading new [23:09] application inside [23:10] your project. Therefore, [23:11] er, the team is very [23:12] similar to the one we [23:13] were doing for launch [23:14] the entire project. You [23:15] Do you remember how we were then? [23:17] started the Django admin project [23:19] momentum trader. Uh, team [23:21] to launch the application- [23:22] start app. We will launch [23:24] portfolio. You will see that [23:25] here automatically [23:26] a heap is generated [23:27] files. And then we [23:29] let's do the same for [23:30] trading. Okay, you see, [23:33] what is the app for [23:33] trade and application [23:35] for portfolio [23:36] created. The last thing [23:39] we need to do, [23:40] this is uh, register [23:41] these applications in the project [23:43] Django. So, we created [23:45] all files for everyone [23:46] application, but for now [23:47] the project is not really [23:48] knows that these files [23:49] exist. So he doesn't [23:51] knows how to read them or [23:52] refer to them [23:53] in any way. Therefore, [23:54] in every Django project [23:55] is this extremely [23:57] the important file settings.py. [23:59] And this is essentially the file here [24:00] , obviously, was [24:01] automatically [24:02] created by Django during [24:04] project download [24:05] . He can [24:06] to turn into [24:07] a very long wall [24:08] configuration if you [24:10] You don't know what you're doing. [24:11] So we'll simplify it. [24:12] We just need [24:13] edit one [24:14] the list here, which [24:15] is called [24:16] installed applications, [24:17] and we just need [24:18] add two attachments, [24:19] which we just created [24:20] . So, we added trading [24:22] and portfolio. Okay, we [24:26] let's keep this. Hmm, this is [24:29] approximately in [24:29] a smart place to [24:31] we could [24:31] commit this and [24:32] send to GitHub. [24:34] One thing we [24:35] we want to do before [24:36] this is to add a file [24:37] .gitignore. So this is the file [24:39] configuration if you [24:40] not familiar, which one [24:41] Git tells you which [24:42] ignore files under [24:44] GitHub submission time [24:45] . I don't want to commit. [24:46] this SQLite file, [24:47] because it is a base [24:48] data, and its trace [24:49] to store only [24:50] locally. So, we [24:51] add star.sqlite 3, and you [24:55] you will see that now [24:56] he must be [24:57] unallocated, and this [24:58] means that he is not [24:59] will be prepared [24:59] for commit. Perfectly. [25:01] Okay, so I [25:02] I am going to do ` [25:04] get_add`. We will prepare [25:05] all for the commit. And we [25:06] we are going [25:07] commit this to GitHub. [25:08] I will say `initial' [25:10] project download [25:12] and registration [25:14] Django application. [25:17] Perfectly. So, you [25:18] you can see [25:19] summary of all changes [25:20] here. We added 25 [25:21] files, added 23 [25:22] lines of code and all that [25:23] will now be on GitHub. [25:25] Okay, so now we [25:26] let's move on to the section [25:27] this project, where we [25:28] let's start modeling [25:29] data in our database [25:30] data for our [25:31] project [25:31] algorithmic [25:32] trade. Hmm, you see, [25:34] I opened trading/models.py, and [25:36] This is where we will create [25:37] all our base models [25:39] data. Therefore, [25:40] specifically, we [25:41] let's create the following: [25:43] stock. Um, for now I [25:47] I'll just add it here. [25:48] pass operators. We [25:50] let's create another one [25:51] model called um [25:53] price data. You can [25:56] to see that all these [25:57] Python classes [25:57] inherited from [25:58] Django model class, which [25:59] essentially means that his [26:00] need to be compared with [26:02] database table. [26:03] Um, and then we [26:04] momentum score needed, [26:05] trading signal and rebalance event. [26:15] Um, momentum score, trading signal [26:24] and rebalance event. Okay, this is, [26:28] um, each of these [26:29] classes corresponds [26:30] database tables. AND [26:32] then in these classes [26:33] we need [26:34] actually indicate which [26:35] um fields exist in this [26:36] tables, setting [26:38] attributes for the class. [26:41] For example, if you have [26:42] there's a stock ticker, um, and this [26:43] character field, you, [26:45] probably will do [26:46] something like that. I am now [26:49] I will get the exact names. [26:50] fields um. Okay, so, [26:53] regarding shares, we [26:54] you will need uh ticker. [26:56] We will need [26:57] ticker. I don't know what. [27:00] gave me this [27:00] autocomplete. Uh, [27:02] we will need [27:03] name. To us [27:04] a sector will be needed. [27:07] Sector. Uh, us [27:12] market will be needed [27:14] capitalization. To us [27:18] will need active [27:19] indicator. So, this is [27:21] will allow us [27:21] disable the action in [27:22] in the future if [27:23] We don't need it. [27:25] We will need a date. [27:26] creation and date [27:27] renewal. Good. [27:29] So, for each of [27:31] these fields to us [27:31] type must be specified [27:33] fields, and Django has [27:34] built-in classes for [27:35] each of them. Therefore, [27:37] the ticker will be [27:37] character field. And I [27:44] I'll comment on them for now. [27:46] so that they don't give out [27:47] errors in the file. Therefore, [27:49] the ticker will be [27:49] character field. [27:51] Its maximum [27:52] the length will be 10. It [27:53] it will be unique that [27:54] equals truth so that [27:55] we didn't have [27:56] duplicate tickers in [27:57] database. And we [27:58] we are going to add [27:59] before that base index [28:00] data so that the search is [28:01] fast. The name will be [28:03] similar. This is symbolic. [28:04] field. Its [28:06] maximum length, [28:08] let's say 255. And, uh, we [28:10] let him be [28:11] empty during [28:13] creation. Sector ( [28:15] Sector) will also be [28:16] character field. [28:22] Its maximum [28:24] the length will be 100. And [28:25] it will also be blank [28:26] is true. Market [28:29] Market cap [28:33] will be great [28:34] integer field. [28:38] And it will be null- [28:39] permissible. And we [28:42] let's also let him [28:43] to be empty. For [28:46] those who are not familiar, [28:48] null is a restriction on [28:49] database levels, and [28:50] then blank allows [28:51] you on different levels [28:52] in Django too [28:53] install this [28:54] value to false. Therefore, [28:57] uh, Django has a built-in [28:59] panel [28:59] administrator, where are you [29:00] you can log in and [29:01] view different [29:02] database connections. [29:04] And setting blank [29:05] equals true allows [29:06] you actually [29:07] send empty [29:08] values ​​in these forms. « [29:11] Active, how are you? [29:12] can you imagine, [29:13] will be a logical field. [29:17] And we are going [29:17] set for him [29:18] value for [29:19] default "true", [29:20] so that when you create [29:21] one of them, it [29:22] virtually automatically [29:23] became active. Its [29:24] will not be necessary [29:25] activate manually [29:26] later. "Created in" [29:28] will be a date field and [29:29] time. And then, uh, Django [29:32] has this built-in [29:34] assistant named [29:35] auto_now_add. And this means, [29:37] what if these objects [29:37] are recorded in the database [29:38] data, Django [29:39] automatically [29:39] sets their time [29:40] creation, uh, without [29:41] necessity [29:42] to indicate it [29:43] manually. So, we [29:44] let's say that auto_now_add [29:46] equals "true". And, uh, [29:47] this shouldn't be [29:48] capital letter. Here [29:50] Yes. And then, " [29:51] updated in" will be very [29:52] similar to "created in" [29:53] , except that [29:55] instead of auto_now_add it will be [29:56] just auto_now. Okay, this is it. [29:57] first database table [29:58] data that we [29:59] we are going to create. [30:01] To actually [30:02] create a table [30:03] databases in the database [30:04] data, you need [30:05] generate file [30:06] Django migrations, and then [30:07] move your base [30:08] data. So, uh, [30:09] let's do this. [30:11] Team for [30:12] creating migration [30:14] Django-python manage.py makemigrations. [30:17] So, as you can see, this is [30:18] creates three new bases [30:20] data or, sorry, [30:21] four five. My [30:22] counting today [30:23] not successful. Five new [30:24] database tables, [30:25] one for each, [30:26] which we just created [30:27] here. And when we [30:29] we migrate, you will see [30:31] that the initial [30:32] migration was [30:33] applied, and all these [30:34] tables now [30:35] exist in our database [30:36] SQL ID. So, I just [30:40] I'll copy, uh, some [30:41] configurations for [30:42] other tables to [30:44] to start. So, the data [30:46] about prices will have these [30:48] fields. Uh, they have [30:49] foreign key for [30:51] shares. They have [30:52] date field. They have [30:55] opening price, [30:56] maximum, minimum, [30:57] closure and volume, [30:58] adjusted price [30:59] closure and created [31:00] in. Momentum score has the following [31:06] fields: Promotion, date [31:07] calculation, momentum score, [31:09] rank, quintile, [31:10] upper quintile, [31:11] beginning of the period, [31:12] end of period and [31:14] created in Last, [31:17] which one are we going to [31:18] to create, [31:19] called a trading signal. [31:20] This one is a little more complicated, [31:22] because he has [31:22] value [31:23] enumeration. AND [31:25] some base indices [31:26] data. So, I'll insert [31:28] this is here. We can do it. [31:30] review. Therefore, [31:31] trading signal [31:32] includes this list [31:33] types of signals. And, by [31:34] essentially, what is it for? [31:35] is used, you [31:36] you can see [31:37] link to it [31:38] here. This type of signal [31:39] has limited choice, [31:41] which means that any [31:42] what is the meaning, [31:43] written in this line [31:44] database, there should be [31:45] one of these three. AND [31:48] they are set up here [31:49] as two tuples, so [31:50] that the first meaning is [31:51] what actually [31:52] is recorded in the database [31:53] data, and the second [31:54] the meaning is that [31:55] will be displayed in [31:56] in a readable form [31:57] person, in different [31:58] parts of the interior [32:00] Django codebase. [32:01] So, here's what it is. THERE ARE [32:04] -eh, if you walk by [32:05] for the rest of the model, it [32:06] has a foreign key [32:07] for shares, date [32:08] signal, signal type, [32:09] foreign key for [32:10] momentum estimates, [32:11] target quantity, [32:13] target value, [32:14] field of reason, [32:15] is being done, [32:15] is performed on that date [32:17] creation. Then in [32:18] you have this metafield that [32:19] allows you to add [32:20] additional [32:20] configuration to [32:21] database tables in [32:22] Django. So, she has [32:24] database table, [32:26] we specially [32:26] wrote the name [32:27] tables for her. Uh, [32:29] it has order by [32:30] default when [32:31] you are extracting data from [32:32] database, and then [32:33] she has these two [32:34] database indexes. [32:36] The last, uh, the last [32:37] database table in [32:39] trading application [32:40] is called an event [32:41] rebalancing. Therefore, [32:42] if we add this [32:44] here, uh, these are all the fields, [32:45] which she has. Date, [32:46] total number [32:47] analyzed [32:48] shares generated [32:48] buy signals, [32:49] generated signals [32:50] sale. So, these are the goals [32:51] numbers, this is the number [32:52] how much [32:52] signals of each [32:53] type was created. [32:54] Total cost [32:55] portfolio, status [32:56] implementation. This is also [32:57] has a choice limitation. [32:59] Uh, a message about [33:00] error, then date [33:01] creation and date [33:02] completion. So, what [33:03] we need to do [33:04] now, this [33:04] regenerate our [33:05] migration files. However [33:07] , uh, we can't [33:08] to do it, uh, again, [33:10] because we contributed [33:11] too many important things [33:12] changes to these tables [33:13] database for this [33:14] goals. So, uh, what am I [33:17] I'm going to do this. [33:18] just delete this one [33:19] database. We [33:22] we are going [33:23] recycle it. Uh, [33:27] here. And we must have [33:34] the opportunity to perform, [33:36] uh, deletion during [33:37] this initial [33:38] migration. Never [33:40] do this in [33:40] work application, [33:41] but this is the fastest [33:42] way to fix this [33:43] here. We will fulfill [33:45] migrations python manage.py make. [33:47] And then we will execute [33:49] python manage.py migrate. OK, [33:50] perfectly. Now everything [33:52] looks good. Ahem, [33:53] let's do it again [33:54] Let's start our server. [33:57] While I'm thinking about it, [33:58] you sometimes see that I [34:00] I run pm as [34:01] alias for python [34:02] manage.py. So, it's simple. [34:05] It becomes like this when I [34:06] I'm launching it. Me [34:08] just a little faster [34:10] to write because I [34:11] I do this all the time. [34:12] So, we execute pm [34:13] run server and look at [34:14] application, and everything is still [34:16] looks great. Therefore [34:17] , uh, after we [34:19] we did it, we can [34:20] go to our [34:21] another Django application and [34:22] create several there [34:23] database tables. [34:28] Okay, now let's go. [34:29] let's create a few [34:30] data models for [34:31] our portfolio- [34:32] application. First model [34:35] called a portfolio. We [34:38] let's do the same thing as [34:39] last time, just [34:41] let's create them all, and [34:42] then we will fill in [34:43] later. Next [34:44] is called position, [34:45] that is, she [34:46] represents something in [34:47] your brokerage [34:48] account you are [34:48] actually own. [34:52] Next-trade. So, this is [34:56] represents an agreement, [34:57] which you are sending [34:58] to the broker to change [34:59] positions in your [35:00] portfolio. And the last one- [35:02] this is an indicator [35:03] productivity. Therefore, [35:13] like last time, I [35:16] I'll just copy everything. [35:17] indicators, and we [35:18] we can them [35:19] to discuss. So, here, [35:21] in your briefcase, you have [35:23] name, description. They [35:25] quite simple. [35:26] Initial cash [35:27] funds and current [35:28] cash, [35:29] total cost. It [35:30] all decimal fields, [35:30] because they [35:31] intended for [35:32] currency representation. [35:33] Identifier [35:33] SnapTrade user and [35:34] identifier [35:35] account [35:36] SnapTrade. So, these [35:37] briefcases will be [35:38] specifically related to [35:39] SnapTrade partner. AND [35:40] then the secret [35:41] SnapTrade user [35:42] will be a field, [35:42] security-related [35:43] which we need [35:44] save. Active, [35:45] created by and [35:46] updated about. Many [35:47] similar things to [35:48] some of our others [35:49] data models. We [35:50] we are also going [35:51] place here [35:52] metaclass. So, we [35:53] we are going to provide [35:54] this is a certain table [35:55] database called [35:56] briefcases, certain [35:57] order. Next [36:01] data model [36:02] is called a position [36:03] position. So, she [36:04] has a foreign key to [36:06] the portfolio you [36:07] just created here. [36:08] She also has [36:08] foreign key to [36:09] shares that [36:10] are in [36:10] trading model. You [36:11] you will see what is here [36:12] wavy line. It [36:13] because this name is not [36:14] defined. So, there are two [36:16] ways to fix this. [36:17] You can [36:18] import it here [36:20] so: with trading.models import stock. [36:23] That will be enough. [36:25] Or you can [36:26] pass it on as [36:27] line and specify the name [36:28] Django applications and [36:29] model. That's what I'm here for. [36:31] I decide to do it. We [36:32] just use [36:34] trading.stock, and that's it [36:35] will fix it. So, for [36:37] Do you have any positions? [36:37] number, average [36:38] cost, current [36:39] price, current [36:40] value, [36:41] unrealized [36:41] profit/loss, [36:42] unrealized [36:43] profit/loss in [36:44] percentage, date [36:45] last update [36:46] and creation. We [36:47] also add [36:48] metaclass to this [36:49] database models. [36:50] So, we will give this [36:52] database tables [36:53] special name [36:54] positions. We will do this, [36:55] so that the portfolio and the stock [36:56] could only have one [36:58] position. So, this is [36:59] unique constraint [37:00] together. And then we [37:01] let's put this in order [37:02] by default, where [37:03] it will be [37:03] sorted by [37:04] current decrease [37:05] value. So, if [37:06] you will pull them out [37:07] database, then [37:08] your default [37:09] position with the largest [37:10] current value [37:11] will be displayed [37:12] above. Let's [37:13] let's move on to the model [37:14] trade. This option [37:15] starts with [37:16] several lists [37:17] for selection constraints, [37:18] as we discussed [37:19] earlier. So, uh, in [37:22] we have types of agreements and [37:24] status options. Uh, [37:25] we have an external [37:26] key to the briefcase, [37:27] foreign key to [37:28] shares. We will do it here. [37:30] the same as before. THERE ARE [37:35] -uh, we have a type of agreement, [37:36] which, uh, depends [37:38] from the list of types [37:39] agreements. So, that's essentially it. [37:40] does not depend on [37:41] is this a purchase agreement, [37:43] or sale. Then in [37:44] we have quantity, price, [37:45] number of completed, [37:47] executed price, [37:48] order value, [37:49] status. It depends. [37:50] from the list [37:51] status options [37:52] here. So, it is possible [37:54] status options for [37:55] agreements: pending, submitted, [37:57] partially completed [37:58] done, canceled [37:59] or rejected. Then [38:01] we have an external [38:02] order identifier [38:03] , identifier [38:04] SnapTrade orders, [38:04] message about [38:05] commission error, [38:06] submitted at, executed at, [38:07] created by Fr. I just [38:08] I'll take a minute, uh, [38:09] explaining the difference [38:10] between the external [38:10] identifier [38:11] warrants and [38:11] identifier [38:12] SnapTrade orders. When you [38:13] send the deal [38:14] via SnapTrade to Alpaca, [38:15] Alpaca assigns her [38:16] order identifier [38:17] . This is what [38:18] assigned in the field [38:19] external [38:20] identifier [38:20] warrant here. And then [38:21] SnapTrade also assigns [38:23] him an identifier [38:23] warrants, and we [38:24] we keep both in [38:25] database on our [38:26] side. The reason why we [38:27] called it external [38:28] identifier [38:29] warrants, not Alpaca, [38:30] is that you [38:31] could [38:32] use that [38:33] modeling itself [38:33] data for trading [38:34] also in another [38:35] broker. So, perhaps, [38:36] could you take this [38:38] system, split it [38:39] and set it up so that [38:40] so that she can bargain [38:41] both on Alpaca and on [38:42] Interactive Brokers. Both [38:43] identifiers [38:44] orders of these brokers [38:45] will be stored in [38:46] this field [38:46] external [38:47] identifier [38:48] warrants. So, let's [38:49] move on. We [38:50] also add [38:51] metaclass to this [38:52] data models. We have [38:54] database table with [38:55] the name of the auction, [38:56] sorted by [38:57] descending, created in [38:58] . So, when you have them [39:00] pull out the last ones [39:01] will be on top. AND [39:02] then two indices [39:03] database, index [39:04] unification by [39:04] briefcase and [39:05] status, and then [39:06] index of association by [39:07] status and created [39:08] in. So, it's simple. [39:09] question [39:10] productivity. [39:11] Latest base model [39:12] data in this application [39:13] is called a metric [39:14] productivity. I [39:15] I'm going to add all [39:16] fields here, and we will [39:17] Let's consider. So, in [39:19] we have a foreign key [39:20] for the portfolio, we have [39:21] date, general [39:22] value, monetary [39:23] cost, cost [39:24] positions, daily [39:25] income, total income [39:26] , total number [39:27] deals, winning deals, [39:28] unprofitable transactions and [39:29] created by Fr. And then, [39:30] this won't surprise you [39:32] at this stage, we [39:32] we are going to add [39:34] metaclass. So, we [39:35] we are going [39:36] assign to this [39:36] database table [39:37] indicators [39:38] efficiency, [39:38] unique constraint [39:39] for the common [39:39] use on [39:40] portfolio and date. [39:41] So, essentially, you don't [39:42] you can have an indicator [39:43] efficiency twice in [39:44] the same day [39:45] for a specific portfolio. [39:46] Of course? And then, [39:47] order by [39:48] default here [39:49] will be in descending order [39:50] date, and there is one index [39:52] databases. So, this is [39:53] there was a lot [39:53] data modeling. [39:55] The last thing we need [39:55] need to do this [39:57] create migration [39:58] Django to do all this [39:58] applied. Therefore, [40:00] Do you remember that? [40:01] the command for this is python [40:03] manage.py make migrations. It [40:05] will create a file. Therefore, [40:06] you see that [40:07] model created [40:08] briefcase, model [40:09] indicators [40:09] efficiency model [40:10] positions and model [40:11] trade. This is exactly what [40:12] what did we expect. AND [40:13] then we can [40:14] apply this file [40:16] migration by [40:16] using python manage.py [40:18] migrate. And boom, that's all for [40:19] our modeling [40:20] data. Now we [40:21] we can commit this. [40:22] So, I'm going to [40:23] add all our [40:24] indexed files [40:25] . We will see that [40:26] indexed if [40:27] you will run get status. [40:28] There should be a whole [40:29] a bunch of good files. [40:30] And then Oh, actually, [40:31] if you look at [40:32] these are, uh, some of these [40:33] things we don't [40:34] we want [40:34] index. [40:35] Pycache files are required [40:37] add to our get [40:38] ignore. So let's [40:40] Let's do it now. And, uh- [40:42] Well, I think we [40:46] we want to do this [40:47] . Pycache. And then, if I [40:50] I'll come here and get it. [40:52] all things from Pycache files [40:53] . This file. This file. [41:02] That's it. Hmm. I think [41:06] which is actually better [41:07] way to do it, I [41:08] I think GitHub has Uh, [41:16] yes, that's what we need [41:17] necessary. So, GitHub [41:20] has a repository under [41:21] called gitignore, and they [41:23] provide, uh, templates [41:24] gitignore by [41:25] default for [41:26] different types [41:27] software projects. [41:28] So, this is gitignore for [41:29] default Python, and [41:31] I'll just add it. [41:32] to all content [41:33] our gitignore below, [41:34] seems. So, now, [41:36] if we move on to [41:37] our control [41:38] versions, you will see, [41:39] that we will reject this [41:40] change and add ours [41:41] gitignore change. All files [41:42] pycache now [41:43] are excluded. Therefore, [41:44] if we get [41:45] status, yes, just that, [41:46] that we edited. [41:48] Um, so we'll do [41:49] commit and, say, [41:51] source files [41:52] data modeling and [41:53] migration. Perfectly. [41:56] Submit this to GitHub. [41:58] Okay, now we [41:59] let's set up our [42:00] panel [42:00] Django administrator, [42:01] which will help us [42:02] to manage everyone [42:03] different models [42:04] the data we just [42:05] created. The first thing that [42:06] we need to do [42:07] for this, this [42:07] create [42:08] superuser for [42:09] our application, [42:10] who can actually [42:11] enter [42:11] administrator [42:12] Django panel. Therefore, [42:14] the command for this is python [42:16] manage.py create super uh create [42:18] superuser. So, this is [42:21] will offer us [42:22] just pass on [42:23] some information. [42:24] So, the name [42:25] The user will be Nick. [42:26] Um, address [42:27] email not [42:28] it matters, so I [42:29] just enter nick@test.com. [42:30] I will enter the password here. AND [42:32] I will confirm later. [42:33] your password. Steeply. [42:35] So, now we can [42:36] open [42:37] administrator [42:38] Django panel, which is [42:38] one of mine [42:39] favorite parts of the web [42:40] Django framework. Therefore, [42:41] we will run python [42:43] manage.py runserver. She [42:44] just launches our [42:45] standard server [42:46] developments. If you [42:47] open this in [42:48] browser, [42:49] move this to [42:50] screen so you can do this [42:51] to see, and then, uh, [42:52] just add {slash} [42:54] admin at the end of, uh, this [42:55] URLs, enter [42:56] the credentials you [42:57] just sent to [42:58] Django, and boom, you're in [42:59] administrator [43:01] Django panels. Therefore, [43:02] now here for real [43:03] little information, [43:04] except for some [43:04] information about groups [43:05] and users. It [43:06] what we want [43:07] change. We want [43:08] add all models [43:09] the data we just [43:10] created, to [43:10] administrator [43:11] Django panels. I am now [43:12] I'll show you how. [43:13] make. You will see, [43:15] that in two applications, [43:16] which we created, [43:17] application for [43:17] portfolio and applications [43:19] for trade, each of [43:20] This file admin.py has them. [43:22] And we want [43:22] edit it like this, [43:24] so that more of our [43:25] data models [43:26] were reflected in this [43:27] administrative [43:28] Django panels. Therefore, [43:30] Let's start with the portfolio. [43:32] We want to create [43:33] four classes. [43:34] Administration [43:35] portfolio. In essence, we [43:36] let's create one class [43:37] for each created [43:39] we data models. [43:40] Administration [43:41] portfolio, and this has [43:42] to inherit from [43:44] admin.modeladmin. And then you [43:47] Do you want to decorate this? [43:49] using admin.register, [43:50] and then Django models, [43:53] which you are trying [43:54] register. Therefore, [43:55] if you keep this, [43:56] we will get this [43:57] error on our [43:58] development server, which [43:59] says that the portfolio is not [44:00] defined, therefore [44:00] let's import [44:01] this is now. From portfolio.models [44:03] import portfolio. [44:06] Save this, and our [44:07] the development server has [44:09] boot. If [44:10] we will move on to [44:11] administrative [44:12] panels, we now [44:13] we can see it. THERE ARE- [44:14] uh, where did it go? [44:15] Here it is. Okay, now. [44:17] you can see what is [44:18] administration [44:19] portfolio. Uh, and here [44:21] there is nothing, because ours [44:22] the database is empty, [44:23] but you can add [44:24] portfolio, [44:25] review yours [44:26] portfolio, [44:26] list them, etc. [44:27] Uh, and just like you [44:29] do...Excuse me, yes [44:30] just like you do it, [44:31] you can add very [44:33] similar [44:33] administrators for [44:34] the other three models [44:36] data in this project [44:37] . I will show you what I am. [44:38] I mean. Therefore, [44:42] if you just [44:42] copy this three times, and [44:44] then in the first [44:45] in case you change [44:46] portfolio per position, [44:47] in the second portfolio [44:48] for trade, and then in [44:49] third-portfolio [44:51] on the metric [44:51] productivity. Uh, [44:58] you can see that in [44:59] our file has three [45:00] mistakes because we are still [45:01] did not import them into [45:03] this file. So let's [45:04] Let's change this. Position, [45:06] trade and metrics [45:07] productivity. [45:08] Okay, now if we [45:10] let's move on to [45:10] Django administrator, eh [45:18] -hey, he's asking me [45:19] log in again. I [45:21] I'll move this to the bottom of the screen. [45:22] which I am writing down. [45:26] Okay, you can. [45:27] see all four [45:29] Django models. Therefore, [45:30] again, here for now [45:31] that there is nothing, therefore [45:32] that our database [45:33] empty, but we [45:33] we want to check [45:34] this is later. Hmm, the only thing, [45:35] What else I would add is that [45:37] what can be done in this class [45:38] to convey a lot [45:39] really powerful [45:40] functions to make [45:41] its customizable. [45:42] For example, when you [45:43] for the first time you will go to [45:44] portfolio page in [45:46] administrative [45:47] Django panels, you [45:48] you will see a list of all [45:49] your portfolios, and you [45:50] can you specify which [45:51] fields actually [45:52] will be displayed in [45:53] this list, [45:54] passing the attribute [45:55] list display, [45:56] like this. Similarly, [45:58] you can [45:59] filter and [46:00] search on this [46:00] page, page [46:01] portfolio list. AND [46:03] these two attributes [46:03] determine what you are after [46:05] filter and by what [46:06] you can search. AND [46:07] then you can too [46:08] specify some fields [46:09] read-only, so [46:10] people could not enter [46:11] and edit them. [46:12] So, you indicate [46:13] read-only fields [46:14] like a list, like this. I would [46:15] said, like a motorcade, here it is [46:17] Yes. I just... [46:18] I'm going to improve. [46:20] the rest of our models [46:21] data, like this. And this [46:24] our new [46:24] administrative [46:25] page. So, we [46:28] let's move on to [46:29] trading application [46:30] , and I will also copy [46:32] the desired type there [46:33] final version [46:34] this page. And we [46:35] we can go to [46:36] beginning of the actual [46:37] integration with one of [46:38] data providers [46:39] in this project. [46:40] Perfectly. So, we [46:41] let's just spend [46:42] a little check [46:43] working capacity in [46:44] Django admins. [46:46] Let's see how it goes. [46:48] looks. So, you [46:49] you can see that [46:49] each of us [46:50] Django applications have [46:51] your own section. [46:52] There is a portfolio section and [46:53] trade section. If [46:55] you want to see [46:56] only one section, [46:56] just click on [46:57] him, and he will show [46:58] only this section. AND [47:00] then all our different [47:01] database models [47:02] registered here. [47:03] So, this is exactly what we [47:05] we want. We can [47:06] commit this [47:07] now. We'll do, uh, [47:09] we get, uh, period [47:10] . Oh, there's more" Why this one [47:12] Is the guy still here? "I [47:15] I think we just [47:16] we want to take it away [47:18] and move. AND [47:19] then we will commit [47:21] it. Configuration [47:25] Django administrator. [47:28] Steeply. Download this [47:30] on GitHub. And now we [47:31] we can go to [47:32] integration with [47:33] data provider [47:34] Massive. Okay, now we [47:35] let's move on to [47:35] creating our [47:36] integration with Massive, [47:37] which is [47:37] data provider [47:38] stock market, [47:39] which one will we be [47:40] use for [47:41] obtaining data about [47:42] stock prices for [47:42] of this Momentum project. [47:43] Usually when you [47:45] integrate with [47:46] third-party API or [47:47] system, this new [47:48] integration [47:49] placed in a folder [47:50] called services. Therefore, [47:51] you will see that I already [47:53] set up in [47:53] trading application [47:55] a directory called [47:56] services. And inside [47:57] this catalog we [47:58] we will place a new one [47:59] a file called massive [48:01] client.py. Now [48:01] make sure you [48:03] added this file init.py, [48:04] so that other Python code in [48:05] this codebase could [48:07] read it. [48:08] Inside this [48:09] we will add a new file [48:11] a class called massive [48:12] API client. And we will begin [48:15] very simple, simple [48:17] specifying the init method, [48:18] which will accept self. [48:20] He will accept the key. [48:21] API, which can be either [48:23] optional [48:24] line, that is, it will be [48:25] a string object that [48:26] by default has [48:27] value none. And then [48:30] we will also indicate [48:31] number of requests per [48:32] minute to am, which is [48:33] whole number and by [48:34] default [48:35] equals five. Therefore, [48:39] this is because Massive has [48:40] speed limit. [48:41] On the free [48:42] you can level [48:43] send only five [48:43] requests per minute, and [48:44] we must [48:45] build this [48:45] system in such a way that [48:46] to follow [48:47] this restriction [48:48] speed so that it doesn't [48:49] disrupted work [48:49] our program. Hmm. [48:51] The first thing we [48:52] need to do this [48:53] to set up, uh, [48:54] some logic for [48:56] reading our [48:57] API key. So, I [48:58] I'll paste it from here. Uh-huh [49:00] , and I'm going to add [49:01] some are missing [49:02] imports here. So, this is [49:07] V1 of the Massive API client. I [49:09] I am going to explain that [49:11] does all this. So, uh- [49:12] uh, this REST client [49:13] comes with [49:14] the Massive Python package, which [49:16] we need to add [49:17] to requirements.txt. So, we [49:18] add it, and then [49:20] we are launching ours again [49:21] pip install. We, uh, [49:22] sorry, we want [49:23] to be in the catalog [49:24] project one. And then [49:26] we perform UV pip [49:28] install -r requirements.txt. [49:31] Perfectly. So, this is [49:32] brought us Massive One, [49:33] which we indicated, and [49:34] then some [49:35] dependencies next to [49:36] him. So, urllib, websockets [49:38] and certified. Hmm. Next, [49:46] what we want to do, [49:47] is to start creating [49:49] big bulky [49:50] a method called fetch [49:51] stock data. And this will allow [49:53] we receive data [49:54] about prices from massive [49:56] APIs. It looks like this. Oh, I [49:58] clearly inserted the wrong thing [50:00] need. Get data [50:02] about stocks. That's it. [50:04] So, what is missing here? [50:05] some imports. To us [50:06] need to import [50:07] list and dict. So, both [50:09] they exist in the package [50:10] typing. That's it. Good, [50:14] let's see, [50:15] What does this, uh, take? [50:17] this fetch stock method, or, [50:18] sorry, fetch stock method [50:20] data that we create. [50:22] So, he accepts [50:23] ticker, start date, [50:24] end date, [50:25] multiplier, [50:25] adjusted limit [50:26] period of time, [50:27] cache usage, [50:28] maximum [50:29] number of attempts and [50:30] repeat delay. There [50:32] a lot of things. Therefore, [50:37] let's start with the very [50:38] simple. We want [50:40] return aggregated [50:42] data about stocks. It, [50:44] probably the simplest [50:46] way to do it. We [50:50] let's start with the list [50:51] empty eggs, and [50:52] then we sort through [50:54] massive list of eggs [50:55] client and add them [50:56] all like dictionaries to [50:57] this list, first [50:59] than to return them from [51:00] functions. There is a bunch here. [51:03] of unspecified things, [51:04] which we need [51:05] to fix. Therefore, [51:07] The first is this one. [51:08] period of time or, [51:09] sorry, date-time is [51:11] undefined variable [51:12] . We just need [51:14] import it. It [51:15] built-in library [51:17] Python. From date-time [51:19] import date-time. [51:22] The following is the key [51:24] cache. Um, and uh, we [51:27] We are implementing it now. [51:30] So, every time we [51:31] send an API request [51:32] to Massive because it [51:34] has strict restrictions [51:35] speed five [51:36] requests per minute, [51:37] we never want [51:38] send duplicates [51:40] requests. Because of this we [51:41] we will cache [51:42] requests during their [51:43] sending. And if we [51:45] we are going [51:45] realize [51:46] caching, us [51:47] must be specified [51:48] cache key. This is how we [51:49] We will do it. Every time, [51:50] when we send [51:51] ticker request from [51:52] a specific start date [51:54] and completion date, [51:55] we will keep [51:56] it's like a cache key, [51:57] just adding it all up [51:58] together like this. Then [52:00] we can update this [52:02] new method [52:02] data acquisition [52:03] shares to ask [52:05] this cache key is like this. [52:07] Good. The last thing we need [52:09] it is necessary to specify that [52:11] logger. Hmm, that's very [52:13] universal logger [52:14] Python, which can [52:15] realize [52:16] practically in any [52:17] which Python project. [52:19] He looks like this. [52:20] Okay, next thing [52:22] we need to do, [52:23] is to find a way is [52:24] protest. Hmm, [52:26] good way fast [52:27] sort server [52:28] things in Django projects- [52:29] to use it [52:30] the so-called team [52:32] Django management. [52:32] Control command [52:33] Django is, in essence, [52:34] just a CLI that you [52:35] you can run it, and [52:36] which automatically [52:37] will read the whole [52:38] the last code from your [52:39] codebase and something [52:40] will do. Therefore, [52:41] let's set up [52:42] one. Um, teams [52:44] Django management always [52:45] are in the folder [52:46] under the name management/commands [52:48] . So, we will create this [52:49] folder. We will create [52:52] init.py file. Hmm, this is [52:58] need to be placed in [52:59] command. Yes, no, it is in [53:04] the right place. Um, [53:05] Okay, one more thing we need [53:06] need to do this [53:08] easy to set up [53:09] empty command [53:10] control from which we [53:11] we can perform [53:12] iterations. So, here we are [53:13] Let's create a new file. [53:15] It will be called [53:20] pull massive data.py. And I [53:24] inserted this. I just [53:26] asked by ChatGPT [53:27] write to me [53:27] empty command [53:28] Django management. Therefore, [53:30] you run them like this [53:31] just like we launched [53:32] all other Django commands [53:33] in this project. You [53:35] run python manage.py, [53:36] and then the team name [53:38] management. So, in [53:39] in this case you would [53:41] executed python manage.py [53:42] pulling massive [53:44] data. If you [53:45] look what this is [53:46] does, he has [53:47] some arguments. Uh, [53:48] we are actually now [53:49] let's delete these [53:50] arguments, because they [53:51] we don't need it. Uh, and [53:52] we will also remove [53:53] help command, line help uh [53:55] . I expect that this [53:56] just say "hello" [53:57] world "when we do this [53:58] let's start. [53:59] Let's see. Perfectly. [54:02] So, that's right. [54:03] configured. We [54:04] can we update this [54:05] control command, [54:06] so that it actually [54:06] pulled out some [54:07] data from Massive. Therefore, [54:08] let's do from [54:10] trading.services.massiveclient import [54:14] Massive API client. Uh, yeah [54:17] then here we will say, [54:18] that client equals Massive [54:20] API client. And then [54:21] We will take the client out. [54:26] Good. Let's see what [54:28] now it will happen. Settings [54:30] does not have the Massive attribute [54:32] API key. So, if you [54:34] Look here, I [54:35] said that we can [54:37] or pass the API key [54:38] as an argument for this [54:42] service, or, if it [54:43] null, then it will try [54:44] read it from [54:45] our settings [54:46] Django. So, our file [54:48] settings.py if you have it [54:49] open and manage [54:50] F here for Massive, there [54:52] there is nothing, and this [54:53] because we haven't seen it yet [54:54] set up. What are we [54:56] Let's do it now, okay? [54:58] let's set up, um, we [54:59] let's configure, um, the package [55:00] Python, which will allow [55:02] it is easy for us to read [55:03] environment variables. [55:06] This package [55:07] It's called Python decouple. [55:09] We will add it to [55:11] requirements.txt. Python-decouple. We [55:15] let's reinstall our [55:17] project requirements.uv pip [55:19] install -r requirements.txt. Steeply. [55:22] So, Python decouple now [55:23] installed. Then [55:25] we need [55:25] configure example [55:28] environment variable [55:30] environment. So, um, this is [55:32] will look like this. [55:34] .env.example. And the first [55:39] the record that we [55:40] we are going to introduce, this [55:41] massive API key. Now, um, then [55:43] that we establish [55:44] this variable, uh, in this [55:46] example file [55:47] environment, [55:48] doesn't actually mean [55:49] that our Django project is already [55:51] reads it. To us [55:54] need to be entered [55:55] changes to our file [55:56] settings so that, um, [55:57] force Django project [55:59] read this. I [56:00] I'll show you what I have. [56:01] in mind. So, if [56:03] you will go to [56:04] settings.py is closer to [56:05] beginning, we will add [56:06] this import. From decouple [56:08] import config. This package [56:10] decouple is the one we [56:11] just installed in [56:13] our file [56:14] requirements.txt. Then, [56:15] near the end, you [56:16] you can add this. [56:18] Massive API key [56:19] equals config, and then [56:20] you are just passing on [56:21] variable name [56:22] environment, and everything [56:23] ready for the race. [56:25] Now that we [56:25] we are launching this [56:26] control command, [56:27] she must have [56:28] another mistake. Interesting. [56:30] She has no other [56:31] errors. Um, exchange [56:33] these services. I [56:39] I think it's easy for us [56:40] need to copy [56:42] .env.example in regular [56:44] .env file. Let's [56:46] let's try this now. [56:49] Steeply. So, we have [56:50] configured massive [56:52] API client, massive [56:53] API client object. [56:55] Which one is it now? [56:56] uses this [56:57] API key. Now, [56:58] obviously this API key [57:00] won't work because it [57:01] just random [57:02] line. So I [57:03] I will return to my [57:04] massive portal. [57:08] Now, obviously, this [57:09] integration is not [57:10] will work with this [57:11] API key, hard [57:12] encoded here because [57:13] it's just random [57:14] line of text. [57:15] First of all, I [57:16] need to insert [57:17] my actual API key. [57:18] So let's do it. [57:19] this is now. Let's insert [57:21] its. Then we want [57:22] update our team [57:23] control so that it [57:25] actually pulled out [57:26] some data from stocks. [57:27] So let's [57:28] we'll do, uh, fetch stock [57:30] data. We'll try Apple, and [57:39] I am missing here. [57:40] some arguments, [57:41] but Python will tell us, [57:42] What exactly don't I like? [57:43] enough. So, the date [57:44] start and date [57:45] completion. [57:46] Let's try 2026 05 10 and 2026 [57:53] 05 15. Let's see what it is [57:56] will give us. Good, [58:00] now we get [58:02] error that says [58:03] that our Mass API client [58:04] does not have an attribute cache [58:05] , and that's because I [58:06] copied and pasted [58:07] some things, not [58:08] after reading them [58:09] carefully enough. [58:12] So what do we want? [58:13] do in the init method, [58:14] is to add a simple cache [58:16] in memory, which [58:16] It looks like this. [58:23] Let's see what [58:24] will happen. Good, [58:25] it seems like this [58:26] potentially [58:27] It worked. Uh, us [58:28] print operator is missing [58:30] in the control team [58:31] Django. So, let's say that [58:33] this is the result. [58:35] Let's see which one [58:36] result. And then I [58:39] I'll put it there too. [58:40] a stopping point to [58:41] then it was possible [58:42] play with the data. [58:44] Okay, it looks like we [58:45] received some data from [58:46] Massive API. This is important [58:48] landmark. Our integration [58:50] officially, hmm, it works. [58:53] So, the next thing we [58:59] we want to do this [59:02] add a little logic [59:03] before that, so that [59:04] speed limit [59:05] really followed [59:06] . And for this we [59:07] we are going to make [59:10] a bunch of changes to this [59:11] Massive client. First, [59:12] what we want to do, [59:14] is to add an attribute to [59:15] init method called" [59:16] queries per minute. [59:17] This will look like this [59:18] Yes. Everything we do, [59:19] -it's nothing special [59:21] . He just takes, uh, [59:22] this argument from the method [59:24] init and assigns it [59:25] as a class attribute. [59:27] Then we want [59:28] add, uh, method [59:28] speed limit, [59:30] which is called, [59:31] when we send [59:32] him with ours [59:32] requests. He [59:33] It looks like this. [59:34] So, uh, what's missing here [59:35] some imports that [59:36] causes errors, [59:37] so we are going [59:38] import package [59:42] Python time. And let's [59:46] let's consider what this is [59:47] does. So, he [59:48] gets the current time. [59:51] He is looking for all of us. [59:53] query times that [59:54] older than 1 minute. AND [59:57] if it is longer than [59:58] number of requests per [60:00] a minute than [60:00] stated, he [60:01] goes into mode [60:02] sleep, that's just [60:03] means he is waiting [60:05] a certain amount [60:06] seconds. Uh, and then [60:08] he adds more [60:09] data by query time [60:10] . So, uh, obviously [60:12] What do you notice? [60:13] doing this, this is what [60:14] time of requests not [60:15] determined. We also [60:16] we are going to add [60:17] this attribute in the method [60:18] init. It will be simple. [60:19] a list of all the different ones, [60:21] hmm, this is a list of hours [60:22] marks of each [60:23] once when we [60:24] sent outgoing [60:25] request to Massive API. Good [60:29] . So what can we do? [60:32] do it now, this [60:36] update this method, [60:37] so that he has a bunch [60:38] handlers for [60:42] working with this. So, I [60:45] I'll just paste it here. [60:46] the method needed, and we [60:47] we can walk along [60:48] new logic. So, this is [60:52] new method, and I [60:53] some are missing [60:54] imports, so I them [60:55] I will add. Good. Let's [61:00] Let's walk through it. [61:02] So, this is the final [61:03] method version [61:04] data acquisition. [61:06] The first thing he does is [61:08] this gets the cache key [61:09] for the query, and then [61:10] first checks [61:11] cash on delivery [61:12] data. If in cache [61:14] there is nothing, that is [61:15] the cache is not warm, you would [61:16] they said we need [61:18] get fresh data. [61:19] Then he checks [61:20] our limitations [61:20] speed and [61:21] determines whether it is necessary [61:22] him to move to [61:22] sleep mode. If he [61:23] no need [61:24] to switch to mode [61:25] sleep, then this is the one [61:26] the method we [61:26] used [61:27] earlier. He will try. [61:29] to do it [61:29] maximum [61:30] the number of times that [61:31] specified in the definition [61:32] method, and the value for [61:34] default [61:34] equals three. Therefore, [61:36] in fact, he [61:36] will send outgoing [61:37] request to Massive up to three [61:39] times and [61:39] will try [61:40] get data. And if [61:41] he will fail three times [61:42] in a row, then this [61:43] will actually lead to [61:44] a severe setback, and [61:45] everything will be ready. Therefore, [61:48] This is great. Problem [61:49] is that [61:50] if we are going [61:51] to do the Momentum project and [61:53] we want to send [61:54] outgoing requests for [61:55] a huge amount [61:56] shares, then we [61:57] will need to have [61:58] opportunity to do [61:59] much more than [62:00] just send [62:01] one request for one [62:02] action and wait 5 [62:03] minutes. We need [62:05] to be able to [62:05] process it in batches. [62:07] So, we do this, [62:12] adding this new one [62:13] a method called Fetch [62:15] Bulk Momentum Data. He [62:16] uses different APIs [62:17] from Massive, which [62:18] allow you [62:19] get [62:20] grouped daily [62:21] eggs. So, I'll insert [62:22] this method, and we together [62:23] let's consider what he [62:25] means. Good. Therefore, [62:30] this is a great method. Here [62:33] method signature. [62:34] So, he accepts [62:36] tickers, and then [62:37] accepts date [62:37] calculation. So, this is [62:39] mass sampling from [62:40] using [62:41] grouped [62:41] daily API Massiv. He [62:43] sends only two [62:44] API calls and receives [62:45] all tickers for 12- [62:46] monthly and 1- [62:47] lunar days [62:48] simultaneously. Therefore, [62:49] let's take a look [62:50] it. The first thing he [62:52] does, it checks, [62:53] or the settlement date [62:54] is equal to zero. If [62:55] she is not there, then he [62:56] assumes that we [62:56] we count on [62:57] today. Then he [62:58] gets date 12- [62:59] a month ago, [63:00] he gets the date 1- [63:01] a month ago. You [63:02] Do you remember when I [63:04] described this project [63:05] at the beginning that we [63:06] we use [63:07] momentum 12 minus 1. This is [63:08] means that we [63:09] considering momentum [63:11] stock prices for the last [63:12] 12 months, for [63:13] except for the last 30 [63:14] days. So, they [63:15] they call it 12 minus [63:16] 1. For this reason we [63:17] need to get [63:18] 12th lunar day [63:18] antiquity, we [63:19] need to get [63:20] 1 month old day [63:20] prescription. Then, for [63:22] all tickers in the list [63:23] tickers that we [63:24] provided, created [63:27] this is a reflection where [63:28] we say that we have [63:29] no price 12 months [63:30] therefore, we do not have a price [63:31] 1 month ago. Then [63:32] we are checking our [63:33] speed limit, [63:35] sending the first one [63:36] API call. We [63:36] send a request to [63:38] Massive endpoint [63:39] Group Daily Aggs, and maybe [63:40] will be useful [63:41] review [63:42] documentation on this [63:43] question. Okay, that's it. [63:49] documentation for [63:50] the endpoint that [63:51] we are going [63:52] to use to [63:56] get all our [63:57] price data for this [63:58] trading moment [63:59] strategies. So, you [64:00] you can see what's there [64:02] written to receive [64:03] daily OHLC data, which [64:04] means volume [64:05] opening/closing [64:06] at the maximum/ [64:06] minimum, and VWAP, [64:07] weighted average for [64:08] volume price, for [64:09] all US stocks on [64:10] a certain trading day. [64:12] This endpoint [64:13] returns the full [64:13] market coverage in [64:14] one request that [64:15] allows to conduct [64:16] massive [64:17] analysis, processing [64:18] big data and [64:19] research of a wide range [64:20] market [64:20] efficiency. Hmm, [64:22] just to repeat, [64:23] the reason why we [64:24] we use this [64:25] end point, [64:26] is that [64:27] free Massive API [64:29] has limitations [64:29] speed in one [64:30] request or, sorry, [64:32] five requests for [64:33] a minute, and we want [64:34] to trade widely [64:35] spectrum of stocks, therefore [64:36] one request for [64:37] just a minute [64:38] not enough. Hmm, [64:40] sorry, five [64:40] requests per minute [64:42] not enough. So we [64:43] we are going to unite [64:44] this is in requests to this [64:45] endpoint. [64:47] Let's dig deeper. [64:48] and let's create it now. [64:49] Hmm, or excuse me, [64:50] let's go deeper [64:51] and let's check this out [64:53] now, I would say. [64:54] If you come here [64:56] and say" fetch, er, bulk [64:59] momentum data ", and then we [65:05] let's restart [65:06] this control command [65:07] Django, what do we get? [65:10] Two are needed if [65:11] specified is incorrect [65:13] argument. So, this is [65:14] takes tickers and date [65:15] calculation. Let's [65:16] let's take Apple and Nvidia and [65:24] let's see what it is for us [65:25] will give. For the date [65:28] we calculate [65:29] Let's take the 10th number. [65:32] Steeply. [65:35] Unsupported type. [65:37] Oh, that's an expected date. [65:39] So we will do from [65:41] datetime import date. And we [65:44] Let's make a date. Steeply. [65:51] Try this. He has [65:54] there is no fetch attribute. [65:57] So, here is a backup [65:58] the method that I [65:59] will need to be added. [66:02] It is called [66:08] backup method [66:09] receiving mass [66:10] momentum data. We [66:12] we are going to him [66:13] add, and I will explain that [66:14] he does. So, he [66:18] returns to [66:19] the old method, if [66:20] Group API [66:21] daily request not [66:22] works. Let's [66:23] let's find out why we have [66:25] A failure occurred. So, this is [66:28] was here, reserve [66:29] data processing method [66:30] momentum. AT, [66:38] backup method [66:39] processing of missing [66:40] data was also [66:41] missing. So, there is [66:45] all kinds of special [66:46] the logic that we [66:47] adding now, and me [66:48] will have to do this [66:49] to figure it out. Therefore, [66:51] backup method [66:51] processing of missing [66:53] data tries [66:54] find the nearest ones [66:55] dates. So, obviously, [66:59] if we ask [67:00] stock market data [67:01] over the weekend, this [67:02] will encounter some [67:03] strange problems. [67:04] So, we have a few [67:05] shifts of days. He [67:06] will go for 3 days in [67:07] future or for 7 days [67:08] in the past. And this [67:09] helped to solve [67:10] many problems that [67:11] we could have done it differently [67:12] see in this API. [67:14] If we run this [67:16] now, this API, [67:17] obviously, handles [67:18] a bunch of data. Ago [67:19] return requests [67:20] may take some time [67:21] time. Uh, and I think that [67:22] this is what it is now [67:23] is happening. He [67:24] just gets data [67:26] from Massive. Let's give it to him. [67:27] a minute to process. [67:28] Okay, I clicked. [67:29] pause button to [67:30] just understand that [67:32] is happening. It [67:33] somehow funny. I [67:36] wrote it down as Nvidia, and [67:37] not like a ticker, and it [67:38] was causing the problem. [67:41] So, the next thing we need to do is [67:42] need to do this [67:44] just add this one [67:44] method of obtaining [67:45] several shares, and we [67:46] we can start [67:47] management team [67:48] and see what [67:49] will return from Massive. [67:52] Steeply. So, our [67:54] management team [67:55] has a result. [67:56] Let's see, [67:57] what is the result? Therefore, [67:59] we have the price of Apple 12 [68:00] months ago and price [68:01] Apple 1 month ago. [68:03] Similarly for Nvidia, in [68:04] we have a price for 12 months [68:05] ago and price 1 month [68:06] ago. Both of these prices [68:07] actually quite [68:08] performed well [68:09] over the last 12 months. [68:10] So, our integration [68:11] Massive is located in [68:12] pretty good [68:13] condition. In the next [68:14] section we will move on [68:15] to build our [68:16] integration of fast [68:17] trade. Before we [68:19] let's move on to [68:20] Snap Trade integration, I [68:21] forgot one thing: us [68:22] you just need [68:25] record all our [68:26] changes. So, we have [68:27] This PyCache file [68:28] constantly [68:29] returns. I [68:30] I am going to delete. [68:31] its. And then everything else [68:36] looks good. We [68:37] we want to preserve our [68:38] changes here, here, here. [68:40] So, I'm going to do everything [68:42] to place it. Uh-huh, [68:45] This looks good. [68:47] Good. Um, MVP of the Mass [68:51] integration. Steeply. [68:58] Then, uh, in the same [68:59] catalog where we [69:00] created a massive [69:01] client, we will create [69:03] new file named [69:04] Snap Trade Client. In this [69:09] files, uh, we [69:10] let's create our client [69:11] The Snap Trade API, which I [69:12] conveniently called trading [69:14] executor. Because it is in [69:15] mainly that, for [69:16] what will we do? [69:16] use. [69:18] Sorry. As before, [69:20] we are going to add [69:21] new init method. He is not [69:23] will have no [69:23] arguments, so this [69:25] It will be very simple. Ahem, [69:26] and we are going [69:27] configure Snap Trade SDK [69:29] here. He will look [69:31] like this. And as soon as I did that [69:32] I'll insert it, you'll see, [69:33] what do we need [69:34] to do a few things. [69:35] First, we [69:36] need to add [69:37] some are missing [69:38] imports from django.com. [69:40] Import settings. I [69:42] not really [69:43] spent time on [69:44] explanation of this in [69:45] this video, but, according to [69:46] In fact, we saw [69:47] before we have this one [69:49] settings.py file. This [69:50] import allows [69:52] import everything that [69:53] defined in settings.py, in [69:55] any other folder in [69:56] your application [69:58] Django. So, that's the first one. [70:03] The second thing we need [70:04] need to do this [70:05] install SnapTrade SDK [70:08] Python in our project. Therefore [70:11] , I will show you [70:12] project page on [70:14] PyPI. She looks like this. [70:18] Yes. So, this is the name [70:20] the package we want [70:21] add to requirements.txt. AND [70:23] then we can [70:24] import it into [70:25] this file. So, we [70:27] let's move on to [70:28] requirements.txt, we will add [70:29] it. And then, as [70:31] before, we will perform [70:33] UV pip install-r requirements.txt. [70:35] So, this establishes [70:36] many dependencies, [70:38] which are [70:38] subdependencies [70:39] SnapTrade Python SDK. And then [70:44] here, this allows us [70:45] import SnapTrade Python [70:47] SDK is like this. So, from [70:52] snaptrade.client import SnapTrade. [70:54] Okay, uh, we [70:55] we are going to add [70:56] a whole bunch of different ones [70:58] methods to this SD to [70:59] this, uh, client [71:00] SnapTrade now. One of [71:02] they will be called [71:03] synchronization [71:04] portfolio permissions. THERE ARE [71:05] -oh, what am I going to do? [71:06] to do, uh, for this [71:07] integration, this is what [71:08] I am going [71:09] to form all this here [71:10] yes, and every method is wise [71:12] will simply call [71:13] error exception not [71:14] implemented during [71:15] settings. Uh, here [71:16] we will need [71:17] import our [71:18] portfolio model. [71:19] So, from portfolio.models import [71:22] portfolio. Uh, us too [71:25] will have to [71:26] import this [71:27] list mind with typing import list [71:29] . So, it's simple. [71:31] type hint with [71:32] built-in [71:33] typing libraries [71:34] Python. And then we also [71:35] will have to [71:36] import model [71:37] positions. So, from [71:39] portfolio.models import uh, import [71:42] position. Steeply. So, this is [71:45] first method. Second [71:48] the method is called [71:49] execution of orders on [71:50] I will buy. So, this is [71:51] the method you [71:51] call as soon as possible [71:53] chose securities, [71:54] that you want to buy, and [71:55] it's time actually [71:56] place a deal. [71:57] So, as before, we [71:58] we are going to perform [71:59] error raise not implemented. [72:01] There are also no [72:02] some imports. Therefore, [72:04] we want [72:05] import from [72:06] trading.models import stock and trade. [72:10] And then, uh, this one [72:11] decimal type [72:12] just take it from [72:13] built-in [72:14] decimal [72:14] Python libraries. Good [72:18] . Next method [72:21] very similar. Instead [72:22] execution of orders on [72:23] we will buy [72:24] we are going to perform [72:25] sell orders. And we [72:27] we are also going [72:28] to supplement this [72:29] error not implemented. [72:31] Cool, it works. Uh, [72:34] then we are going [72:35] create a method for [72:37] updating our [72:38] trading statuses. [72:39] It looks like this. Error [72:41] Raise not implemented. And then [72:46] we need more [72:48] two methods that [72:49] are called get current [72:51] stock price and get available cash for [72:52] trading. I actually [72:54] just insert them and [72:55] I'll explain right now, [72:56] because they are both [72:58] very simple. Before [72:59] I'll do it, we're here. [73:01] two more are missing [73:01] imports. One was [73:03] optional, but [73:04] another is date and time. [73:09] Good. So, for [73:12] receiving the current [73:13] stock prices are very [73:14] simply. We just [73:15] let's take a massive API- [73:17] a client we had [73:19] earlier, and, uh, [73:20] call the method [73:21] getting a price for [73:22] date. This method [73:23] not really [73:24] determined. So, if [73:25] you copy this and [73:26] look here and [73:27] take the test [73:29] action F, it's not there. [73:30] And that's because we [73:31] forgot to add it, [73:32] but it's true [73:32] simple method. He [73:33] It looks like this. [73:39] So, here we have [73:40] link to this [73:41] price obtaining method [73:42] on the date that we also [73:43] didn't add it because I thought [73:45] that he is not [73:45] is used, that [73:46] my mistake. It [73:48] a fairly simple method, [73:49] which we can add [73:50] now. It looks like this [73:53] Yes. So, getting [73:54] prices on the date accepted [73:55] ticker, target date, and [73:56] then the number of days [73:58] admission. Therefore, [73:58] number of days [73:59] tolerance determines, on [74:00] how many days [74:01] massive client [74:02] will go forward or [74:03] back if on this date [74:04] none available [74:05] price data. He [74:07] just causes [74:07] method of obtaining [74:08] data about the action, which [74:09] we created before. [74:10] So, here it really is [74:11] there is none [74:12] custom logic. It [74:13] just another way [74:14] link to this. Uh, and [74:15] the last thing we need here [74:16] is not enough, that's what [74:17] we need [74:17] configure logger [74:18] Python as we did [74:19] earlier. So, we [74:20] we are going [74:21] import [74:22] logging, and then [74:23] logger is equal to [74:24] logging.getlogger like this. [74:28] OK. So, that's all. [74:35] Um, the last thing we need [74:37] need to add this [74:38] method below under [74:39] called um get trading executor uh [74:41] , which returns [74:42] a copy of this. It [74:48] very simple. Looks like [74:50] like this. OK. [74:53] Let's begin. [74:54] create some of [74:55] these methods now. Uh-huh [74:56] , in fact, before [74:57] we will do it, we [74:58] need to be configured [74:59] our credentials [75:00] SnapTrade. So, you [75:00] remember that when we [75:02] first started [75:02] to work on [75:03] large-scale [75:04] integration, we [75:04] had to adjust [75:05] some variables [75:06] environment so that we [75:07] could actually [75:08] connect to SnapTrade [75:09] . Let's do it. [75:10] the same process for [75:12] SnapTrade. So, two [75:13] The variables we need are [75:14] here. I'll just copy. [75:16] these two lines. We [75:17] we can insert them [75:18] here. And then, with [75:20] a small amount [75:22] kung fu like this, we [75:26] we can, uh, [75:27] set them to [75:28] reading from our [75:32] environment, er, file [75:34] env {dot}. Okay, so you [75:36] you can see that in [75:37] we had a massive key [75:38] API equal to [75:39] configuration, and then [75:40] massive API key. We [75:41] did the same for [75:42] secret key [75:43] SnapTrade client and [75:44] identifier [75:45] SnapTrade client. If we [75:47] now let's move on to [75:49] our env {dot} file, and [75:50] then, uh, let's open it [75:52] new SnapTrade window. I [75:56] I'm going to come in here and [75:58] get your... [76:04] Actually, I'm going to [76:05] register a new one [76:07] account [76:08] now. Uh, registration [76:09] , personal, [76:10] continuation. Uh, [76:37] steeply. So, to me [76:39] need to do [76:40] some things here are like this [76:41] as confirmation [76:42] my email [76:43] mail, etc. I [76:44] I'm going to do this in [76:45] separate window. [76:51] They probably are now [76:52] send me [76:53] link for [76:53] confirmation [76:54] email. [76:59] Yes, I got it. [77:03] Cool, my email [77:04] Email confirmed. I [77:05] I think if I now [77:06] I'll update this, it will point to [77:08] it. Yes, now [77:08] generate an API key. [77:11] First of all, I [77:12] need to be configured [77:13] 2FA, so I'm going [77:14] copy this code and [77:16] insert it here. [77:28] Cool, check it out. [77:30] enabled [77:30] two-factor [77:31] shutdown. Great, in [77:33] I have an API key, so I [77:34] I will take it and [77:35] I'll put it like...How do we [77:36] named the variable? [77:38] Identifier [77:39] SnapTrade client. [77:42] Identifier [77:43] client is equal to [77:44] this. And then the key [77:49] consumer, I want [77:50] regenerate it [77:52] and take it like this. AND [77:54] then this is how we [77:55] let's do...SnapTrade...How is it [77:57] is it called? Uh, [77:59] client secret. Oops, [78:03] this is not what I wanted [78:05] take. I wanted to take [78:06] it. Perfectly. Our client [78:12] SnapTrade looks good. [78:16] Let's go now [78:17] let's start creating [78:18] some of these methods. [78:19] So, the first thing we [78:24] we want to do this [78:24] add some [78:25] precautions to [78:26] to make sure that [78:27] the briefcase contains [78:28] certain information that [78:28] we will need for [78:29] its synchronization. [78:30] In particular, we will do [78:32] that. So, we [78:33] let's check if it has [78:33] briefcase and [78:34] identifier [78:35] SnapTrade user, and [78:36] identifier [78:37] account [78:38] SnapTrade. If not, we [78:39] we will cause an error [78:40] value. And then, [78:41] similarly, if we [78:41] we will not pass any [78:42] user secret, [78:43] we will also call [78:44] error error [78:45] value. Then, uh, I [78:47] I will insert all the logic. [78:48] this method, and we [78:50] we can figure it out [78:51] with him step by step. [78:54] Good. So, the next thing, [78:57] what we will do is [78:58] actually send [78:59] API request to SnapTrade to [79:00] get all positions [79:02] account [79:03] user. And then, [79:06] for each position in [79:07] answers about [79:08] positions, we will get [79:09] symbol, number, [79:10] average cost, [79:12] the current price, hmm, we [79:13] let's skip everything that doesn't [79:14] have a symbol or where [79:15] number less than [79:17] zero. And then we [79:19] we will create promotions for [79:20] all of them. After [79:22] creation of shares we [79:23] also create or [79:24] we will update the positions, and [79:25] then we will save them [79:26] all in the database. Therefore [79:28] , that's all for positions. [79:29] Then we'll do something. [79:31] very similar to [79:31] balance sheets. We [79:32] send a request to the API [79:33] SnapTrade under the name " [79:34] get balance [79:35] account [79:36] user ". And then [79:38] , for balance sheet data, [79:39] we will get through all of this [79:41] in the loop and save [79:42] everything is in the database. Therefore [79:43] , we are going [79:44] get balance [79:46] cash from API balances [79:48] and save it to [79:49] current cash in [79:50] briefcases. That's all. [79:52] Now we will move on to [79:53] execution section [79:54] purchase orders. [79:56] The same. I am going [79:57] insert it and [79:58] guide you through [79:59] all of this. A lot of this [80:01] will become clearer, [80:02] when we move on to [80:03] next section [80:04] interface construction [80:05] user. Okay, in [80:07] execution of orders on [80:08] I'll buy the first thing we [80:10] let's do this [80:10] we will make sure that there is [80:11] actual list [80:12] purchases that [80:13] are sent. And if [80:14] he is not there if he is [80:15] empty or its [80:16] no, we just [80:17] return empty [80:18] list. We also have [80:19] similar check [80:20] secret name [80:20] SnapTrade user, as [80:21] we did it in the method [80:22] synchronization [80:23] portfolio positions. AND [80:24] then for each [80:25] stocks on the list [80:25] purchases we will create [80:26] object of trade and [80:27] let's save it in the database [80:28] data. Then we [80:29] we are going to send [80:30] order to SnapTrade, [80:31] using the method [80:32] forced [80:33] placing an order with [80:34] with such a body. So, we [80:35] we indicate the account [80:36] record. This is a warrant for [80:37] I will buy. This is a market [80:39] warrant, that is, we do not [80:40] need to specify [80:41] Here is the price. Time of action [80:43] The order will be valid for 24 hours, and [80:44] notional value [80:45] will be divided into [80:46] shares, which was [80:47] passed to this method [80:49] and stored in [80:49] object trade. The last one, [80:51] what do we need [80:52] indicate, this is a ticker. [80:54] Then we gather. [80:55] get the result [80:56] warrants. We are going [80:57] update the agreement, [80:58] indicating the external [80:59] order identifier [81:00] . We are going [81:00] update the status of the deal, [81:01] so that she would be [81:02] sent. We [81:03] we are going to save [81:04] timestamp when [81:04] she was [81:05] sent, and then [81:06] we are going [81:06] save it in [81:07] executed agreements. [81:08] Order method for [81:09] the sale is very similar. I [81:10] I'll show you now. [81:16] So, we insert [81:17] him here. Starting [81:18] from above, if not [81:19] list for sale, we [81:20] we are going to return [81:21] empty list, and [81:21] if there is no secret [81:22] user, we [81:23] we are going to call [81:24] value error. [81:25] Then we will create [81:26] a variable called " [81:27] executed agreements" and [81:28] we will assign her [81:28] empty list. We [81:29] let's go through all [81:30] current positions and [81:31] let's save them in a set [81:33] . Then, for each [81:34] stocks in the stock list [81:35] for sale, we will start [81:36] cycle. We will miss [81:38] action, if it is not [81:39] is in [81:39] current position. [81:41] So, if I own [81:42] only Apple and Nvidia, and I [81:43] trying to sell a Tesla [81:45] , it's simple [81:45] will miss the Tesla sale [81:47] because she [81:48] not really in ours [81:49] portfolios for sale [81:51] now. That's it. [81:52] these two lines do. [81:55] Then we will try [81:56] create an agreement, [81:57] send the agreement to [81:58] SnapTrade, process [81:58] response to the warrant [81:59] just like we do [82:00] made in the section [82:01] purchase orders, [82:02] while preserving the exterior [82:03] order identifier [82:05] for the deal, keeping [82:06] status of the deal as [82:06] sent, and then [82:07] keeping the mark [82:08] time sent [82:09] warrants. And then we [82:11] let's take all these [82:12] executed agreements and [82:13] We will return them. Here [82:14] there is also some processing [82:15] exceptions. Good, [82:17] Let's move on. IN [82:18] this SnapTrade client has [82:19] another method, first [82:20] than we can [82:21] go to creation [82:22] our interface [82:23] user. It [82:24] status update [82:26] trade. I am going [82:27] treat it the same way, [82:28] by copying it into [82:29] the window you see, [82:31] and I will guide you through [82:32] all of this. Okay, by- [82:33] first. For any [82:36] trade status, [82:37] which we would call [82:40] final, we [82:41] Let's skip them. Therefore, [82:42] if it is done, [82:43] canceled or [82:44] rejected, we just [82:45] return false, that [82:46] means that we don't [82:46] needs to be updated [82:47] its. He is located [82:48] in the terminal [82:49] condition. Then, likewise [82:50] as we checked [82:51] user secret in [82:52] all our others [82:53] methods, if the secret [82:54] user not [82:54] transmitted, we [82:55] we will cause an error [82:56] value. Then we [82:58] we are going to get [82:59] order status for [83:00] using a method in the API [83:02] SnapTrade. Get it [83:03] status and assign [83:04] his old status. [83:05] Excuse me, take it. [83:06] trade status with [83:07] databases and [83:07] appoint him/her [83:08] old status. [83:09] Then take [83:10] trade status with [83:11] SnapTrade and assign [83:12] his object of trade [83:13] in the database. Then we [83:14] we check the pile [83:15] fields, and if they [83:16] installed in [83:17] answers to [83:17] order, we [83:18] we keep them in the agreement [83:19] , and then save [83:20] agreement in the database. [83:21] So, wow, that was [83:22] very fast [83:23] work, um, but this [83:25] SnapTrade API basics. We have [83:27] there is much more [83:27] do in the interface [83:29] user from point [83:30] view of the actual [83:31] connecting our [83:32] accounts [83:33] etc. Hmm, and I'll tell you. [83:35] More about that later. Good [83:38] before we [83:38] let's move on to [83:39] website creation [83:40] and actual [83:41] visible interface [83:42] user for [83:43] of this project, which I [83:44] I know, we are all very [83:45] worries. There are two more. [83:46] services that we [83:47] need to add to [83:48] our system for [83:49] rebalancing and [83:50] calculation of trades. Ahem [83:51] , so we also [83:52] let's place them in this [83:53] services folder. They [83:56] will be called, uh- [83:57] yes, strategic [83:58] mechanism, and they [83:59] will be called, [84:00] calculator [84:01] momentum. So, uh, I [84:05] I know we just [84:06] wrote a lot of code [84:08] for the backend, and all [84:09] glad to see this on [84:10] frontend. Therefore I [84:11] I will try to do this. [84:12] a little bit of [84:13] shorter. First we [84:14] we will consider [84:15] calculator [84:16] momentum. Therefore, [84:19] calculator [84:20] momentum [84:20] will look like this. [84:24] So, I put this in, and I [84:25] I'll just consider it. [84:26] I will explain each method. [84:28] at a high level, as [84:29] they are working. First [84:31] than I will do it, you [84:32] you will see that we [84:33] we get some [84:34] attributes from settings.py that [84:36] we need to specify. [84:37] So it should be 12, and [84:41] it has to be one. And this [84:43] what, uh... the reason [84:44] why are they [84:44] configured as [84:45] variables in settings.py, [84:46] is that [84:47] if you want [84:48] split this project [84:49] and actually accept [84:50] its real [84:51] trading strategy, this [84:52] will allow you [84:53] go from 12 minus [84:54] one momentum to any [84:55] -any other strategy [84:56] momentum that you [84:57] do you want [84:57] use. [84:58] If you want [84:59] make eight minus [85:00] three, for example. I am not [85:01] sure that there is [85:02] a lot of literature, [85:03] what would it be [85:03] confirmed, but [85:04] from now on it is possible [85:05] to adjust. Good. [85:06] So, the calculator [85:07] momentum has a method [85:08] for calculation [85:09] one momentum for [85:10] shares. So, it's simple. [85:12] allows you to do [85:13] 12 minus one momentum [85:14] , getting a bunch of prices from [85:15] databases. He has [85:17] method" get price [85:18] from the database ", which [85:19] allows easy [85:20] find price data [85:24] stocks in the database. AND [85:25] then the method "get" [85:26] price with API "like, [85:27] but instead of [85:28] contact the database [85:29] data, he [85:30] appeals to [85:31] massive API. AND [85:33] finally, we have a method [85:34] "calculation of grades [85:35] momentum "massively, [85:36] who does everything [85:37] necessary for [85:38] calculating grades [85:39] momentum in [85:40] frontend. I will tell you. [85:42] you about it step by step [85:43] step by step. The first thing he [85:44] does, it checks, [85:45] Is there a settlement date? [85:46] If she is not there, he [85:47] assumes that this [85:48] today. Then he [85:49] checks whether it is [85:50] list of stocks. If it [85:51] no, he gets [85:53] all active shares with [85:54] databases. Then he [85:55] goes through or, [85:56] sorry, gets [85:57] big data [85:58] momentum from [85:59] massive API. And then [86:00] for each share in [86:01] of this massive data [86:02] momentum or [86:03] about that for [86:04] of each share in our [86:05] list of shares he [86:06] goes through them, gets [86:08] 12 and one momentum, and [86:09] then calculates here [86:10] momentum assessment and [86:11] stores it in the database [86:12] data. Um, yes. [86:14] auxiliary method for [86:16] calculation of individual [86:17] momentum estimates, and [86:18] there are also a lot of things [86:19] for ranking stocks. [86:20] So, this method [86:21] is called [86:22] by ranking stocks by [86:23] momentum. He [86:24] takes into account the assessment [86:25] momentum and [86:26] calculates quintiles [86:27] for a specific date. [86:28] Actually we don't have [86:29] such auxiliary [86:30] method, so let's [86:31] let's add it now. [86:37] Good. So, if we [86:40] let's move on to the assessment [86:41] momentum, this is the method, [86:42] which we want to add [86:43] . And we need [86:44] import timestamp [86:45] belt so that it [86:46] worked. If we [86:49] scroll down, [86:50] let's consider, [86:51] How does it work? So, according to [86:53] first, we see this [86:54] regularity in [86:55] many places [86:55] this project, but [86:56] he checks whether [86:57] we send the date [86:58] calculation, and if not, [86:59] he accepts [87:00] today's date for [87:01] calculation purposes. [87:02] Then he pulls out [87:03] each copy [87:04] momentum estimates in [87:05] database and [87:06] arranges them by [87:07] falling grades [87:08] momentum. If [87:09] there are no grades, he [87:10] just does the early [87:12] return. If they [87:13] no, then he [87:14] goes through them all and [87:15] calculates quintiles [87:16] like this. Um, okay, [87:18] let's go back to [87:19] calculator service [87:20] momentum. [87:21] Using [87:22] similar logic, we [87:23] we have the opportunity [87:26] get the top [87:27] quintile, get [87:28] bottom quintile, and [87:29] then update our [87:30] universe of stocks. Therefore, [87:31] this is done in [87:32] massive client and [87:33] gets all S&P tickers [87:34] 500. Then there is a certain [87:35] logic for [87:36] filling in price lists [87:37] data. There is a certain logic. [87:38] to receive [87:39] momentum statistics [87:40] . There is a certain logic to [87:41] checks [87:42] calculations [87:43] momentum. And this [87:44] summarizes the service [87:45] calculator [87:46] momentum. Fourth [87:48] and the last service for [87:50] this project [87:50] called a motor [87:52] strategies. I am going [87:53] to make something like, uh, [87:54] a brief overview of this [87:55] now and then we [87:56] we can go to [87:57] interface creation [87:58] user. So, I [87:59] inserted it here. Uh-huh [88:00] , it has a basic [88:02] init method. And one of [88:03] things that you immediately [88:04] notice when [88:05] look at this, this is it, [88:06] that the only argument [88:07] there's a briefcase for that. [88:08] So, the portfolio represents [88:09] a briefcase, [88:10] connected via Snap [88:11] Trade. So, essentially, [88:13] when you apply [88:14] one of these strategies [88:15] momentum, you [88:16] you need to choose [88:17] portfolio, [88:18] connected to Snap Trade. [88:19] Uh, there's a lot of stuff here. [88:22] The first method [88:23] called should rebalance [88:24] (should be rebalanced). [88:26] And, uh, what does that mean? [88:27] or what it does is [88:28] looking for your last [88:29] rebalancing day, [88:31] and then asks: " [88:32] How many days have passed since [88:33] the last one [88:34] rebalancing? Hmm, [88:36] trading strategy [88:37] by momentum it can [88:38] be weekly or [88:40] monthly. And this [88:41] does, it checks, [88:42] Is it already day? [88:42] rebalancing [88:43] during the last [88:44] week or during [88:45] last month. AND [88:46] if so, he will say [88:47] that you don't need [88:49] rebalance. Following [88:53] step is, um, method [88:54] implementation [88:55] rebalancing, which [88:56] actually performs all [88:57] trade, necessary [88:59] for rebalancing. [89:01] So, he ranks [89:02] momentum stocks, [89:03] generates trade [89:04] signals, and then [89:05] performs trade [89:06] signals. Here is a method [89:09] generation of trade [89:10] signals, which [89:11] passes through and [89:11] determines which agreements [89:12] you need to buy. [89:14] So, this is essentially [89:14] recalculation mechanism [89:15] , and then another one [89:16] method that [89:17] is called [89:17] execution of trade [89:18] signals, which [89:19] does all this. [89:20] Finally, there is a method under [89:20] called get strategy performance ( [89:21] receiving [89:22] productivity [89:23] strategies), where you [89:24] you can see [89:24] productivity [89:25] strategies, and then [89:26] verify [89:26] settings [89:27] strategies if [89:28] want to look at [89:29] it. The last method in [89:32] this is called running [89:33] backtest, what is it really? [89:34] just a placeholder. [89:35] One of the homemade [89:36] tasks that I [89:37] I'm going to give people [89:38] this project, this [89:39] open a request for [89:41] removal before this [89:42] repository with good [89:43] backtesting logic, and, uh, [89:44] I would like to see [89:45] everything you have. Therefore, [89:46] This sums up our [89:48] latest services. Further [89:49] we will start with [89:50] interface [89:50] user. Good, [89:52] now we will start with [89:53] interface development [89:54] user. This is a course [89:55] from algorithmic [89:56] trade, not with [89:56] interface development [89:57] user, so this, [89:59] will likely include [90:00] a lot of copying and [90:01] inserts. Uh, but I [90:02] I will do my best. [90:03] to explain [90:03] development principles [90:05] interface [90:05] Django user by [90:06] progress of work. Therefore, [90:08] First of all, I created [90:10] off screen uh [90:11] catalog in the store [90:13] application called [90:14] templates/trading. And then [90:17] I copied a bunch here [90:19] interface templates [90:20] Django. Uh, come on. [90:21] let's look at a few [90:23] examples. Uh, what? [90:24] would it be simple? Hmm, [90:27] maybe delete [90:29] portfolio. So, that [90:31] what do I want to show [90:32] you here, this is this one [90:33] logic. It [90:34] is called language [90:36] Jinja templates. And if [90:37] have you ever [90:38] used React [90:39] or any other [90:40] library [90:41] dynamic [90:42] interface, I just [90:43] I want to show you what [90:45] this is the Django equivalent, and [90:46] it allows you to have [90:47] loops, conditional [90:48] operators such as [90:49] if statements etc., [90:50] inside your HTML- [90:52] templates. This is enough. [90:53] stylishly. Now this [90:54] is considered a little [90:55] old-fashioned, and many [90:57] serious programs [90:58] just choose React. [90:59] But that's exactly what it is here. [91:01] Now, for our project [91:03] actually [91:03] used this, [91:05] we need to add [91:06] urls.py file. We [91:07] let's go to the store [91:08] addition. We will press [91:10] "new file". He [91:11] will be called urls.py. [91:13] And then we need [91:14] insert something [91:15] similar. This is, in essence, [91:17] URL heap mapping [91:18] addresses with a bunch of HTML- [91:20] templates. So, if [91:22] we will run python, [91:23] Let's clean up first. [91:24] my terminal. If we [91:26] now let's run python [91:27] manage.py run server, we [91:29] Let's open it. We [91:31] we will never see each other again [91:32] standard page [91:33] Greetings Django. We [91:34] we will see the real one [91:35] addition. AND [91:39] It turns out that I [91:40] was wrong because [91:41] missed a step. I you [91:43] I'll show you. Um, in [91:44] mainly to us [91:44] need to go to [91:46] Root level URLs.py [91:47] and import them [91:48] also. So, in the application [91:52] Momentum Trader, which is like [91:54] home of everything [91:55] our project, we [91:56] need, mmm, [91:57] insert, um, URL- [91:58] import addresses from [92:00] trade. To us [92:08] need to do this. [92:12] Representation [92:12] trade like Oh, okay. [92:15] We also lack [92:16] some representations. [92:17] A lot is missing. [92:19] So, apart from ours [92:25] file URLs, nam [92:26] also needed [92:27] add a bunch [92:27] Django views. So [92:29] this will be enough [92:30] large file. He has [92:31] 800 lines. This is a pile, mmm, [92:35] a bunch of submissions [92:36] Django, which handles [92:38] life cycle [92:39] responses to the request in [92:45] application. So [92:46] Bear with me while [92:47] I will finish this. We [92:49] we are going [92:50] return to [92:51] momentum_trader/urls. Um, I [92:52] actually used [92:53] wrong here [92:54] syntax. I want [92:55] use this. [93:00] Turn on Cool. And I [93:05] I think that this [93:06] the frontend has [93:06] download now. [93:11] Maybe I [93:12] will have to [93:13] restart the server [93:14] developments. Yes. Perfectly. [93:30] Okay, our app [93:31] is already running, and you [93:32] you can see that this [93:33] approximately similar to that [93:34] , what did it look like [93:35] demonstration that I [93:36] showed you before [93:37] in the video. However, I [93:38] missed a big one [93:39] step, ignoring [93:41] all these representations [93:42] Django. So I am now [93:44] I'll explain what it is. On [93:48] In fact, each of these [93:49] methods–this is a function, [93:51] which processes the request [93:52] browser. So, [93:54] Simply put, [93:55] can you imagine [93:57] Django presentation as [93:58] a function that accepts [93:59] browser request and [94:00] sends a reply [94:02] browser. That's why [94:04] this argument is request, and [94:05] you can see [94:07] below, what are we [94:07] we render something for [94:08] sending back. And in [94:11] render functions you [94:12] you will see that she [94:13] accepts the request, [94:14] accepts which HTML- [94:16] display template, and [94:17] then takes a certain [94:19] context to insert [94:20] in this HTML template. Therefore, [94:23] if you switch to [94:25] dashboard.html, I'll close these [94:28] windows that we don't have now [94:30] we use. If [94:31] you will look at [94:32] dashboard.html, you will see, [94:34] that somewhere there is a field for [94:35] calculation dates. [94:37] Right here. So, in [94:39] Python functions for [94:39] toolbars [94:41] we calculate the date [94:42] calculation, it [94:43] inserted into [94:44] toolbar, [94:45] and then we can [94:46] see it in [94:47] interface [94:47] user. And this, in [94:50] in principle, how [94:51] frontends are working [94:52] Django. We need [94:55] to work on it, [94:56] to be able to see [94:57] some promotions in our [94:58] portfolio, and we [94:59] Let's do this next. First [95:01] than we will continue [95:02] interface development [95:04] user, we [95:04] let's fill our database [95:06] given a bunch of stocks and [95:07] historical data [95:08] about prices, so that in [95:09] interface [95:09] user were [95:10] things we can [95:11] to see that [95:12] to make sure that [95:13] the project works as follows [95:14] we expect. For [95:16] I will add a new one to this. [95:17] control command [95:19] Django called backfill [95:20] data. This is quite long. [95:26] team, and I don't know, [95:27] or its details [95:28] extremely important, [95:29] because she just [95:32] to fill our [95:33] databases. Actually [95:34] it is not related to [95:35] program, but in [95:36] basically she [95:39] fills our database [95:41] given a bunch of shares, and [95:42] also historical [95:43] data about these shares. [95:46] So, put this in [95:47] your project, and then [95:49] run this command: [95:51] python manage.py backfill_data, a [95:52] then you can [95:54] add--update-universe, a [95:55] then--days 420. And, um, us [95:57] also needed [95:58] make sure that this [96:00] method determined that [96:02] we clearly missed [96:03] earlier. Hmm, I didn't think so. [96:07] that he will be [96:08] to be used in [96:09] project, and it is not [96:10] used, [96:12] except for the sake of, um, [96:13] this, and it's just [96:14] hardcoded [96:16] list of some [96:17] S&P 500 tickers. He [96:18] It looks like this. Good, [96:21] so get the tickers [96:22] S&P 500. I think he [96:24] returns, um, 50 [96:25] largest stocks in the S&P [96:27] 500 or something like that. AND [96:28] these are the stocks we [96:29] first we gather [96:31] fill the database. [96:32] So, you can [96:33] to see that we [96:33] we get about 200 [96:34] and some historical [96:35] prices for everyone [96:36] security, and this [96:37] it will take some time to [96:38] to finish because you [96:39] reach the limits [96:40] speed. So [96:41] launch it, and [96:42] then come back, [96:43] when he finishes. [96:44] Okay, so after [96:45] batch completion [96:46] filling and [96:47] repeated [96:48] loading [96:49] interface [96:49] user you [96:50] you will see the button " [96:51] To list [96:52] momentum" in [96:53] upper right [96:54] forged. You need [96:55] will press on it [96:56] , and loading [96:57] will last a long time. It [96:59] may take 3 [97:00] minutes, because [97:01] need to be done [97:02] many calculations " [97:03] under the hood", [97:04] especially when everyone [97:05] data from Massive API [97:06] are loaded into [97:07] your local database [97:08] data for the first time. But [97:10] when will it be done [97:11] , your home [97:12] the page will look like [97:13] like this: you will have [97:14] Top 10 and 10 [97:15] worst stocks [97:15] momentum, all of [97:16] using logic [97:17] , which you wrote yourself [97:18] . So congratulations. Now [97:20] we just need [97:21] to adjust [97:22] functionality [97:23] trade. We created [97:24] a lot of logic on [97:25] server side, [97:26] but nothing yet [97:27] not tested. So [97:28] Let's begin. If you [97:29] go to mine [97:30] trading signals and [97:31] click « [97:32] Generate signals » [97:33] , you will get this [97:34] error. And you will say: [97:36] "Oh, what is this?" Every time, [97:37] when do you get [97:38] interface error [97:39] user, similar [97:40] to this one, from Django, to you [97:41] you need to come here [97:43] and look into your [97:44] terminal to [97:44] to see what it is [97:45] error. So, here [97:46] It is written that we are not [97:47] we can import [97:48] trading with trading, and this [97:49] because we [97:50] imported it from [97:51] wrong programs [97:52] . It should be imported. [97:53] from portfolio.models, not from [97:56] trading.models. So, it was [97:58] just my mistake, [97:59] when I, uh, coded this [98:00] manually. So, I made [98:01] I saved this change. Hmm, [98:02] we need again [98:04] press" [98:04] generate signals ". [98:09] Active portfolios [98:10] not found. First [98:12] create a portfolio. [98:13] So, we clicked "OK" [98:15] ". And, uh, if you [98:17] I wonder where this is. [98:18] because I know that we [98:19] didn't create this [98:20] manually, uh, create [98:22] portfolio. It [98:24] is in the file [98:25] views.py, which we [98:26] inserted into the window [98:27] signal generation. [98:29] So, we need [98:30] generate now [98:31] portfolio. If we [98:32] click in [98:32] interface [98:33] user on the left [98:34] panels on this tab [98:35] portfolios if we [98:36] click here, [98:37] a request will appear for [98:38] creating your [98:39] first portfolio. [98:41] Let's call it, uh, [98:42] free registration [98:44] educational [98:44] Code Camp guide. We [98:47] Click "Next". AND [98:49] then we will choose Alpaca [98:51] for our broker and [98:52] safely [98:52] let's connect via [98:54] SnapTrade. So, I have [98:55] another window [98:55] a pop-up appeared [98:56] window. It looks [98:58] like this. We [98:59] press" [98:59] Continue ". We are back again [99:01] press" [99:01] Continue. "And then I [99:04] already entered Alpaca. AND [99:05] It says that we [99:07] authorize Snap Trade [99:08] for installation [99:09] connection, etc. We [99:10] press" [99:11] Allow ". Boom. Um, [99:15] this is not very nice [99:16] displayed in [99:17] small window, but [99:18] if you like it [99:19] you will open [99:20] you will see that [99:20] The connection is complete. [99:21] We have $100,000. [99:22] in cash, like us [99:23] expected from [99:24] paper connection [99:26] Alpaca, which we [99:26] installed earlier. AND, [99:28] uh, now we can [99:29] view portfolio [99:30] . So, this is a briefcase. We [99:32] we can [99:32] synchronize it, [99:33] we can generate [99:33] signals, we can [99:34] receive trade [99:35] signals, we can [99:36] view our [99:36] momentum indicators [99:37] etc. Let's do it again [99:38] let's come here and [99:39] let's press" [99:39] generate signals ", [99:40] now that we have [99:41] portfolio set up [99:42] . We have another [99:43] one mistake [99:43] signal generation. [99:45] So let's do it again [99:46] let's fix this in [99:47] Django terminals. Good, [99:49] It is written here that we [99:51] trying to get [99:53] access to [99:54] portfolio.get_current_positions, a [99:55] This does not exist. So [99:57] let's add it. [99:58] If you go to [100:00] portfolio, or, sorry, this [100:02] should be in models.py in [100:03] portfolio application. To us [100:04] need to add [100:05] a method called get [100:09] current positions in the portfolio. So [100:12] let's do this. [100:17] There are actually several [100:19] methods for position, which [100:20] we will need for [100:21] this. So let's [100:22] let's add...Okay, so what do I [100:27] just added? I added [100:34] a method called update [100:35] current value, which updates [100:38] current value [100:39] portfolio that I [100:40] surprised. Me [100:41] had to add [100:42] a method called add [100:43] shares. I added a method under [100:44] called remove shares. They [100:46] are on [100:47] wrong model [100:50] Django. Sorry about that. [100:51] They must not go. [100:52] to the portfolio. They must [100:54] go to position. So, update [100:55] current value, add shares and remove [100:57] shares in position. In our portfolio [101:00] you only need to add [101:01] two very short and [101:02] simple Python methods. [101:05] We need to add [101:06] calculate total value. Everything that [101:08] he does, this [101:09] summarizes the current [101:10] the meaning of all [101:11] positions in the portfolio. [101:12] And then we need [101:13] add a method under [101:14] called get current positions, [101:16] which simply filters [101:17] all positions in the database [101:18] data and selects them. [101:20] So, after all [101:21] we can do this [101:21] return to [101:22] our interface. Em [101:24] , this is not the right window. It [101:25] the window in which we [101:26] have failed. So [101:27] I will refresh the page. AND [101:29] then I press " [101:30] generate signals ". [101:31] I click "OK". AND [101:32] then we are successful [101:33] generated 10 [101:34] trading signals, 10 [101:35] purchases and none [101:36] sale. So, what is in [101:37] us here? We have some [101:39] purchases of Exxon Mobil, RTX, this [101:44] Raytheon? I think. Nvidia, [101:47] Johnson Johnson, Merck, Broadcom, [101:49] Google and Google. I think that [101:51] if you really [101:52] were gathering [101:53] turn it into [101:54] production and [101:55] build on this [101:56] real trade [101:57] system, you probably [101:58] you want [101:59] deduplicate between [102:00] two Google. It, [102:01] obviously one and the same [102:02] the company itself. AND [102:03] then you have AMD and Intel. [102:05] So AMD actually has [102:07] highest score [102:08] impulse, therefore [102:09] let's buy it at [102:10] 100 dollars. Click [102:12] here. Um, we're going. [102:13] choose this portfolio. [102:14] This is the only one we have [102:16] connected. Hmm, and then [102:17] we are going [102:18] make a purchase. [102:19] Okay, so it's written. [102:20] that the purchase order [102:22] successfully submitted. Now [102:23] we want to go to [102:24] Alpaca and see what [102:25] happened. So, this is Alpaca. [102:30] We've seen this web- [102:31] site earlier in the video. [102:33] Um, I'm going to sign in. [102:34] by means of [102:35] account, [102:36] which I created earlier [102:37] . I need [102:39] add your code [102:40] authentication. [102:43] Okay, I am now. [102:44] entered. Um, I get it. [102:48] welcome [102:48] message. What about me? [102:53] surprising, but look [102:54] where Intel actually [102:55] bargained correctly [102:57] . So, we have already grown up. [102:58] by 0.03%, and just like that we [103:00] we can walk and [103:01] to trade the rest [103:02] momentum strategies. [103:04] Um, that's the whole project. [103:06] Thank you guys for that. [103:07] followed us all the way [103:08] time. Um, if you [103:09] want to contact [103:10] me with questions or [103:11] with something else that you [103:12] could do about [103:13] project to [103:14] to learn a little [103:15] more, please [103:15] Let me know. IN [103:16] otherwise, uh, no [103:18] feel free to play [103:19] with this. I would like [103:20] to see how you are, [103:20] Guys, get on with it. [103:21] this project and create [103:22] with his help [103:23] more things or others [103:24] cool things for [103:25] using Massive Trade or [103:26] Snap Trade. And, uh, [103:27] send me any [103:28] What are your jobs? I would [103:29] wanted to see them. [103:30] Thank you.