Error in read_dataset parameters

[Zimovik007]

dftrain = read_dataset("datasets/dataset_initial/dataset_{:02}.json", 0, 13) // to 13 inclusive
y_train = dftrain.pop('round_winner').map(lambda s: 0 if s == "CT" else 1)

dfeval = read_dataset("datasets/dataset_initial/dataset_{:02}.json", 13, 17) // from 13 inclusive
y_eval = dfeval.pop('round_winner').map(lambda s: 0 if s == "CT" else 1)
print(dftrain.head())

I think some train data fell into the dfeval, because:

def read_dataset(template, start_idx, end_idx):
    frames = [ pd.read_json(f) for f in [template.format(i) for i in range(start_idx, end_idx+1)] ]
    return pd.concat(frames, ignore_index = True)

I really have a strong influence on the end result

[kristoff3r]

Hi,

That is a pretty serious issue during training, thanks for highlighting it. I'll fix the template, hopefully it hasn't affected too many submissions.