Cryptocurrency-predicting RNN intro – Deep Learning w/ Python, TensorFlow and Keras p.8

Welcome to part 8 of the Deep Learning with Python, Keras, and Tensorflow series. In this tutorial, we’re going to work on using a recurrent neural network to …


  1. You are a major positive influence in a world of fake gurus.

    Python was the first language I had properly learnt (and still learning).
    Also, since data science was what sucked me into python, Pandas was one of the first few packages I had started using.

    Hence everytime you use 'list(map(fn…' on a series, I wonder, 'why not just .apply' ?

    Love the videos by the way.

  2. To all those getting this error: ValueError: columns overlap but no suffix specified: Index(['BTC-USD_close', 'BTC-USD_volume'], dtype='object')

    USE : main_df = main_df.merge(df,left_index=True,right_index=True)

    INSTEAD OF : main_df = main_df.join(df)

  3. Can anyone comment if this will work for multiple coulmms? Let's say for sports, you have multiple columns for each sport category in basket ball (points, rebounds, assists..) – will the RNN perform succesfully?

  4. One of the very few videos of a guy who knows what he is doing. There are so many videos out there of machine learning noobs, making a trillion mistakes in each and every of their strategies. You got the basics of machine learning right, which was a pleasure to see.

  5. df = pd.read_csv(dataset, names=['time','low','hight','open',f'{ratio}_closed',f'{ratio}_volume'],index_col=0,usecols=[0,4,5])

    A better way to take the columns.

  6. I am trying to implement RNN for batch control. I have 1 input and 4 outputs.
    All of them are in 50 batches each of length 600 each.
    So input A (temperature) has data of 50 batches with 600 values for each of the 50 batches.
    The outputs B,C,D also have the same dimension.

    Could you tell me how do I go about preparing appropriate shape/structure of this dataset to implement RNN?

  7. hi when i run the code instead of numbers in the close and future columns i just get NaN. Could you please help me out?
    LTC-USD_close future
    5777.000000 NaN NaN
    5777.770020 NaN NaN
    5778.009766 NaN NaN
    5778.220215 NaN NaN
    5790.009766 NaN NaN
    5792.930176 NaN NaN
    5795.669922 NaN NaN
    5800.000000 NaN NaN
    5807.000000 NaN NaN
    5809.399902 NaN NaN

  8. Hey, sir. I followed all the codes you have written. But in the end, I got the accuracy of 0.8777 at the first epoch. But yours is 0.5136. This is so wired. I cannot figure out the reason myself.

  9. probably not better, but it allows you to change the columns easily by dropping them using the list.

    names = ["time", "low", "high", "open", "close", "volume"]

    ratios = ["BCH-USD", "BTC-USD", "ETH-USD", "LTC-USD"]

    x = 0

    df_dict = {}

    for ratio in ratios:

    df = pd.read_csv(f"{ratio}.csv", names = names).drop(columns = names[1:4])

    df_dict[f"dataframe{x}"] = df

    x += 1

    if x > len(ratios) – 1:



    df_dict["dataframe0"] = pd.merge(df_dict["dataframe0"], df, on='time', how='left', suffixes = (ratio, ratios[x]))

    main_df = df_dict["dataframe0"]

Leave a Reply

Your email address will not be published.


7 + = 14