Erros when download and preprocess earthquakes and citibikes

[junwoopark92]

• neural_stpp/data/download_and_preprocess_earthquakes.py

  Line 27 in 5029fa0

  df = df[df["place"].str.contains("Japan")]

In 'place' columns of 1990_2019.csv, there are NAN values, so it causes ValueError: Cannot mask with non-boolean array containing NA / NaN values

I think we need to drop rows with NAN values before running the above code.

df.dropna(subset=['place'], axis=0, inplace=True)

However, I'm not sure if this preprocessed dataset is entirely identical to the one that the authors used for evaluation in their paper.

• neural_stpp/data/download_and_preprocess_citibike.py

  Line 31 in 5029fa0

  stds = df.std(0)

this code tries to get std for the float-type columns. so we need to fix this line as below.

df.std(numeric_only=True)

I would like the authors to review these issues.