Python tool to download data of financial market such as price, candales, volume from BINANCE and computing indicators. Long terme project is to build a neural network to spot patterns over a large amount of data and build a AI trading agent with tensortrade.
8h 1m 55s logged
DevLog #22 : I keep on working on redefining the structure of the code. All the sqlite3 operations are now performed by the class MarketLabDataBase. Redefining the structure af the code will take a long time (8 hours for one step) but it is one of the final steps of the project and it will hopefully improve the complexity grade of my project. Thanks for reading 😆!
3h 32m 53s logged
DevLog #22 : I started working on redefining the structure of the code to strengthen robustness and clarify the architecture. I asked Codex to make me a check list of all the modifications to operate. Thanks for reading 😁 PS : Special thanks to @Aaravkatiyar for liking and commenting DevLog #21 and following ! (You are the GOAT!)
8h 1m 6s logged
DevLog #21 : New functionality! You can now download data from a specific date range. I added this feature to prevent long downloads ( ~ 5 minutes for the whole history with a time frame of 15m and 15 times longer for a timeframe of 1m ). I am now working on computing series with incompatible indicators with general_application. Thanks for reading ✌️! PS : Special thanks to @alyaanmughal for liking DevLog #18, @Aaravkatiyar for liking DevLog #20 and @Sreeni for following me !
3h 0m 16s logged
DevLog #20 : New design ! I redesign the main menu by moving 🧩 Manage Custom Indicators and Events, 🗑️ Delete a Column and 📁 Open or Create a Database in the ⚙️ Settings. Small change but it improves the user experience and the understanding of the main menu. Thanks for reading 😄!
2h 0m 11s logged
DevLog #19 : New functionality ! You can now compute series of the same indicator but with different windows (EMA 20; EMA 30; EMA 40 …). It only work for the functions compatible with general_application for now but I am working to extend this feature to the custom indicator functions. Thanks for reading 🙌!
4h 10m 8s logged
DevLog #18 : New indicator and new event ! The filter_savgol applies a Savitzky-Golay filter (filter that works on polynomial regression) and peak_detection. This led me to add a system that allows the use of indicators and events functions that do not comply with the functions general_application and general_event_application (functions that simplify the création of custom indicators and events). The next steps are the packaging and the writing of the README. Thanks for reading 🙌!
7h 50m 23s logged
DevLog #17 : New functionality ! You can now create a file containing your own custom indicators and events. I also created a function called general_event_application which simplifies the creation of new events. Thanks for reading 😄 PS : Special thanks to @bali0531 for liking DevLog #16 !
3h 24m 4s logged
DevLog #16 : New functionality ! You can now calculate statistics of an event to show up in function of two parameters. In The example, the event is a cross between the close and the SMA 100 in function of the volume and the the volume of the quote. I am coming to the end of the project, the next steps are to allowing the user create its own indicators and events. Thanks for reading 🙌 ! PS : Special thanks to @MathBty and @dhruvalya for liking DevLog #15 !
2h 19m 16s logged
DevLog #15 : New functionality ! You can now search a file already created to make returning to work easier on an existing file. Thank for reading 😄 ! PS : Special thanks to @Gerafftes for liking DevLog #14 !
1h 52m 34s logged
DevLog #14 : New fonctionality ! You can now plot a Status with indicators directly with the interface. In the example, Close, SMA 100, SMA 50 and Cross(SMA 100, SMA 50) are plotted on the same graph. The next step is to calculate Stats and plot it too. Thanks for reading 🫡
5h 17m 2s logged
DevLog #13 : New functionalities ! You can now plot any indicators already calculated directly with the interface. The next step is to plot status (like “cross”) and stats. I also added 4 functional indicators in the indicator dict of the interface (ATR, VWMA, Amplitude and Relative). The Indicator Dict is able to contain recommended parameters for each indicator and they are printed when entering parameters to calculate indicators. Thanks for reading 🫡. PS : Special thanks to @Jinan (again!), @CoolDuck and @MathBty for liking DevLog #12, @PAPER_01 for liking my first DevLog. Thanks again to @jinan for the amazing comment about DevLog #12 and @yeshwingadde for following !
2h 54m 47s logged
DevLog #12 : Bugs fixed ! I added in the code verifications for each user entry to prevent incorrect type of entries and reduce program crash. I also added the option Continue after an action. Hope you like the new logo. Thanks for reading 🙌! PS: Special thanks to @shi , @Jinan and @f for liking DevLog #11. Thank again to @f for the nice comment about DevLog #11!
4h 5m 10s logged
DevLog #11 : New functionality ! You can now calculate indicators directly with the interface and delete it. I also added an History to remember the actions that have been done. Hope you like the new design with emojis. I am still taking ideas from you. Thanks for reading 😁
1h 11m 41s logged
DevLog #10 : New functionality ! I have started to build a terminal-based user interface to make Market Lab much easier to use. Here is a preview of the first functionality : Download Data to download every candles of the selected symbol and interval. I am also continuing to work on the heat map feature (look at DevLog #9). If you have advice about the design or whatever of the interface, I will be very happy to listen to it. Thanks for reading 🫶 !
1h 5m 41s logged
DevLog #9 : New functionality ! The probability for an event occurring can now be calculated and plotted. In the example, the probability of a crossover/crossunder based on trading volume is shown in red or in blue. The next step is to calculate the probability of an event to show up based on two variables and visualize the results using heatmaps. Thanks for reading 🙌. PS : Special thanks to @jinan , @logannavaia and @Epicninja3 for liking the DevLog #8. Thanks again to @Epicninja3 and @logannavaia for following this project !
1h 36m 26s logged
DevLog #8 : New event ! The event “under_over” allows us to detect a moment in time where a variable is under or over another. Here the example shows the moment when the SMA 50 is either under or over the SMA 100. I keep on working on the function that allows to compute the probability of an event to show up based on different variables. Thanks for reading 🤠 PS : special thanks to @f for liking the DevLog #7 and the very encouraging comment !
25m 49s logged
DevLog #7 : Following the DevLog #6. Triangular labels can now be plotted on the graph at the location of each event. In the example, the event is the crossover between two EMAs. The next step is to compute the probability for an event to show up based on different variables such as the volume, the number of trades etc. Thanks for reading ✌️ ! PS : special thanks to @benvelzul for following this project !
1h 16m 55s logged
DevLog #6 : New event ! The event « cross » allows us to detect a moment in time when two values cross each other. Here, the example shows the moment when two SMAs cross one another. This event is shown by the second graph which ocilliates between 1 and -1 for a crossover and a crossunder. The next step is to plot labels directly on the graph for each « cross ». Thanks for reading 😄
18m 21s logged
DevLog #5 : New indicator ! I have juste finished implementing the Commodity Chanel Index (cci). This indicator show how much the deviation of the price form its average (SMA) is important relatively to the mean deviation of the price (MD). The math formula is a bit complexe so I joined it below. The typical price (TP) is the average of the high (H), low (L) and close (C) of the same candle. Thanks for reading this DevLog 🫡
29m 6s logged
DevLog #4 : New indicator ! I have just finished implementing the Stochastic Oscillator (or %K). It is a simple indicator which calculate the “position” of the close of the the last candle over n candles. Here is the math formula: %K = 100 * (Close-Low)/(High-Low) with Close the close of the last candle ; Low the lowest point over n candles ; High the highest point over n candles. The timeframe is still 12h because the calculations are much more faster than those for a 1h timeframe ( logically 12 time faster). Thanks for reading this DevLog 🙌
1h 3m 36s logged
DevLog #3 : I have just finished implementing the RSI calculation. This is the RSI 14 close of the BTC on a 12h timeframe. I have more than 10 indicators now !
36m 38s logged
DevLog #2 : New Indicator ! Here is the ATR 14 of BTC since the 17/08/2017 computed on a 12h time frame. Thanks for reading 😆!
12h 16m 36s logged
DevLog #1 : Very first DevLog! I started Market Lab one week ago and heard about StarDance Challenge one month ago. But I got the idea to participate only yesterday so I have about ten hour that won’t count 😥. I already established the data extraction from BINANCE API. I keep on tracking errors and creating new functions to calculate indicators. Thanks for reading 🫡!
