I'm trying to work with catboost and I've got a problem that I'm really stuck with right now. I have a dataframe with 28 columns, 2 of them are categorical. When the data is numerical there are some even and some fractional numbers, also some 0.00 values that should represent not an empty values but the actual nulls (like 1-1=0).
I'm trying to run this:
train_cl = cb.Pool(data=ret_df.iloc[:580000, :-1], label=ret_df.iloc[:580000, -1], cat_features=cats)
evl_cl = cb.Pool(data=ret_df.iloc[580000:, :-1], label=ret_df.iloc[580000:, -1], cat_features=cats)
But I have this error
---------------------------------------------------------------------------
CatBoostError Traceback (most recent call last)
<ipython-input-112-a515b0ab357b> in <module>
1 train_cl = cb.Pool(data=ret_df.iloc[:580000, :-1], label=ret_df.iloc[:580000, -1], cat_features=cats)
----> 2 evl_cl = cb.Pool(data=ret_df.iloc[580000:, :-1], label=ret_df.iloc[580000:, -1], cat_features=cats)
~\AppData\Local\Programs\Python\Python36\lib\site-packages\catboost\core.py in __init__(self, data, label, cat_features, text_features, embedding_features, column_description, pairs, delimiter, has_header, ignore_csv_quoting, weight, group_id, group_weight, subgroup_id, pairs_weight, baseline, feature_names, thread_count, log_cout, log_cerr)
615 )
616
--> 617 self._init(data, label, cat_features, text_features, embedding_features, pairs, weight, group_id, group_weight, subgroup_id, pairs_weight, baseline, feature_names, thread_count)
618 super(Pool, self).__init__()
619
~\AppData\Local\Programs\Python\Python36\lib\site-packages\catboost\core.py in _init(self, data, label, cat_features, text_features, embedding_features, pairs, weight, group_id, group_weight, subgroup_id, pairs_weight, baseline, feature_names, thread_count)
1081 if label is not None:
1082 self._check_label_type(label)
-> 1083 self._check_label_empty(label)
1084 label = self._label_if_pandas_to_numpy(label)
1085 if len(np.shape(label)) == 1:
~\AppData\Local\Programs\Python\Python36\lib\site-packages\catboost\core.py in _check_label_empty(self, label)
723 """
724 if len(label) == 0:
--> 725 raise CatBoostError("Labels variable is empty.")
726
727 def _check_label_shape(self, label, samples_count):
CatBoostError: Labels variable is empty.
I've googled this trouble, but found nothing. My hypothesis is that there is a problem with 0.00 values, but I do not know how to solve this because I literally can't replace these values with anything. Please, help me!