PyAlgoTrade Tutorial (Part 2) - Backtesting Basics

I'm starting a new channel on AI at / @parttimeai
In this video, we learn the basics of PyAlgoTrade. We first download the last 20 years of historical data for SPY. Then we show the results of buying at the close of the first bar and holding the entire period.

[@parttimelarry]

I'm starting a new channel on AI at youtube.com/@parttimeai Source Code: github.com/hackingthemarkets/pyalgotrade-tutorials Buy Me a Coffee: www.buymeacoffee.com/parttimelarry

[@jack_sparrow1049]

Take this series to advance level... Very excited to learn

[@EdwardTilley]

Great tutorial. Champion man!

[@x.e.b.u]

Quality band - Portishead. It is also a place here in the UK. Thanks again for your videos. Do enjoy them.

[@kruthikngowda6061]

keep creating this type of content larry ,love from india

[@jp21in]

Thank you so much for sharing this content. I’m trading out of a Roth IRA so I could care less about capital gains at all. How should I change my strategy? Should I simply do the monthly process more often? I don’t think there’s any downside to being flagged a pattern day. Trader I may have to deposit more money to allow it, but it shouldn’t cost me anything.

[@dservais1]

Thanks a lot for the free course ;-)

[@anqicheng6069]

hi whats the extension for 13:17?

[@enginbulank1879]

Great video!! Thank you. What I am wondering is, will you be able to sell at Close price? How this lib simulate limit SELL or market SELL?

[@non-randomwalker4579]

Thanks for this great video! I noticed that I get now (at Dec 2021) different Adjusted price from yfinance for 2000-01-03 than you got. Adjusted prices seems to change when time goes (new dividents occured) and I think this might cause the inaccuracy. I mean that the closing prices are adjusted for today, not for the last day in the date range used to pull data. Pls. correct if I have understood this wrongly.

[@TooMuchToThink1]

Do you have any advice of back testing a minute strategy developed on tradingview on this system?

[@parttimelarry]

Doesn't TradingView have a built-in backtester for strategies written in PineScript? I have heard some people say they didn't like it, but haven't taken the time to compare and learn why. As far as backtesting a minute strategy, there is a barfeed of type CSV - in the docs you can see that you can pass in a frequency parameter and you can specify the timeframe. And then you can just code the strategy class to define what you want to do on each bar. gbeced.github.io/pyalgotrade/docs/v0.20/html/barfeed.html#module-pyalgotrade.barfeed.csvfeed I'll see if I have time to put together a shorter timeframe backtest example for a different strategy.

[@TooMuchToThink1]

@@parttimelarry thanks for the quick response! I’ll look into it. I noticed in your videos you pulled yahoo finance data. My understanding is that from yahoo finance I can get o/h/l/c day data, not minute data, but I’ll follow along your video and look into it. My strategy uses multi time frame analysis. But tradingviews built in backtester only goes back about 2 months, not over years and not over multiple stocks (all s&p500 stocks for example)

[@timcain1734]

Another great video. I would be interested in a video that shows how a single python algo program can do backtesting or be run so it trades. For instance, if you run the program with backtest = True, the program will do backtesting. If you run the program with backtest = false, it will be live and trade.

[@juliusfenata]

Hi Larry, if we use the yfinance, how can we get the adjusted price data that include splits & dividends value?

[@HechamelaTamala]

Awesomeness, just what o mese it

[@prestorianer]

in your example the AdjClose of the second day is lower than the first day. everything works fine. But, i started from 1999-04-30 with an AdjClose of 88.3. At the next day (1999-05-03) the price is 89.93. so its higher and the excution fails, since the quantity is to high, too. Is there a way to execute at the calculation day?

[@jameswang6714]

Anyone knows the Python Extension used at 13:17? Thanks ahead.

[@jameswang6714]

@parts Time Lary Thanks, I will contact you.

[@parttimelarry]

Hey this is a spam bot or scammer please don't contact them

[@ziad3274]

@@parttimelarry Yeah please what is the name of that extension ?

[@mila.elnekave12]

Can anyone recommend the best IDE for PyAlgoTrade. i currently use Pycharm, can not write save data to csv. any suggestions?

[@parttimelarry]

I use Visual Studio Code

[@mila.elnekave12]

@@parttimelarry Thank you!

[@brucebane7401]

amazing!!!

[@brucebane7401]

@@parttimelarryl8024 is this the admin of the channel????

[@parttimelarry]

Hey this is some bot, please ignore

[@rverm1000]

i like the way you go about learning new things. if i was better at python it would be somewhat easier.

[@jack_sparrow1049]

can you make a tutorial on coding a custom indicator using pandas and time series data

[@danielrv1325]

This seems great, I have an idea I want to backtest, will try this. Also, which data service would you recommend the most? Nothing crazy, some OHLCV daily data for several years, I am willing to pay but not a crazy bloomberg terminal acc of course haha. Again, great content!!

[@Itsdlu]

Which Python PATH did you run it in on VS code?

[@Itsdlu]

ModuleNotFoundError: No module named 'pyalgotrade'

[@madam_im_adam]

Sweet Portishead tee!

[@rantg]

i am very curious about REAL algorithms, the type of signal processing tools to trade and not moving averages.

[@bartubmaiden]

@parts Time Lary Don't think he got it

[@randomdude79404]

Yes like implementing a strategy out of a research paper (statistics based etc...) Rather than technical analysis ?

[@mr.a1880]

First.

[@theonewhowas7709]

please stop slamming keys