13 comments

  1. 1tczx93 says:

    Hi Sentdex,

    I have been trying to debug my error of Execution Timeout. Changed the time
    period to a shorter one and checked my codes. Still cant find any problem
    with it. Would you mind linking the sample code so I can debug it easier? i
    cant find it in the url you posted.

    Thanks

  2. Kartheyan Sivalingam says:

    Hi,

    Can you please make a video (series) on the networking library Twisted (an
    event driven networking library) – you must of heard of it. It’s a
    networking library which is not part of standard library, the documentation
    for the library is weak and there not many good tutorials on YT, so can you
    please me a series on how to use Twisted. Thanks.

  3. alex preo says:

    unrelated to this tutorial, but is there a way to a) create a moving
    average for a talib function or should i just rewrite the functions for my
    own needs? b) source price data or function outputs for previous days’
    closes?

    any help is very much appreciated and these tutorials are a-mazing

  4. RedstoneIs Awesome says:

    Hey, +sentdex can you make a python module that converts python, to c, and
    then c to assembly.
    I understand assembly, but I want to make an OS, to make it easier to
    create OSes
    Just if your wondering, I know assembly, but not C.
    I have not found a way to directly convert python to assembly.
    Thanks, if you can!

  5. Ryan Albert says:

    Do you have any videos where you talk about sentiment vs trends in the
    stock market and a strategy to help find those?

  6. RedstoneIs Awesome says:

    And, to go along with that, I am making a CPU module(you can get CPU specs).
    I am working on the processor’s getting clock speed.
    When I am done, can you review the module?

  7. martin richardt says:

    Hi,

    I ‘figured out’ a simple strategy that beats the market. Calculate the
    prediction as

    su = sum([p1, p2, p3, p4])
    p= 1 if su == 4 else (-1 if s <=-2 else 0)Everything else stays the same. If all classifiers predict 'buy' (su==4) then buy else if 3 or 4 classifiers predict "sell" then sell.

  8. Luiz Rogerio Blumenthal says:

    Hi,

    if you have an algorithm that is great at being wrong, wont an algorith
    that does exactly the opposite be great at being right? thanks

  9. Bat Man says:

    Hi,

    Thanks for doing this! I really appreciate it! However, I do have a
    question.

    When I was initially running the algorithm the first time. I accidentally
    put elif p == -2: instead of p == -1: However, that error made the backtest
    produce much better result. However, I don’t understand why such phenomena
    occurs?

    Thanks in Advance!

  10. Sida Liu says:

    DEAR Harrison, I must say that I love your tutorial VERY MUCH! They really
    showed me a lot!
    After I watched your tutorial on Recurrent Neural Network, I realized that
    what you were doing here is using 4 algos regardless of sequence. Why not
    using RNN to predict the prices? Hope you can append another tutorial of
    how to use RNN in finance use. Thank you again for sharing your super power
    with us.

Comments are closed.