Pool Creation Error - TypeError: must be real number, not NoneType

[irmakyucel]

Problem: After using a loaded model using catboost load, in the model.predict_proba() function I am facing the error below:
Initially I thought it was about automatic pool creation of catboost so I explicitly created the Pool() but faced the exact same error in that function.

Traceback (most recent call last):
File "/argo/staging/script", line 14, in
model_predict(date)
File "/opt/workflow/app/pipeline/skorlama.py", line 22, in model_predict
pool = Pool(data=df_skor.values, label=None, cat_features=model.get_cat_feature_indices())
File "/usr/local/lib64/python3.9/site-packages/catboost/core.py", line 645, in init
self._init(data, label, cat_features, text_features, embedding_features, pairs, weight,
File "/usr/local/lib64/python3.9/site-packages/catboost/core.py", line 1222, in _init
self._init_pool(data, label, cat_features, text_features, embedding_features, pairs, weight,
File "_catboost.pyx", line 3763, in _catboost._PoolBase._init_pool
File "_catboost.pyx", line 3827, in _catboost._PoolBase._init_pool
File "_catboost.pyx", line 3736, in _catboost._PoolBase._init_objects_order_layout_pool
File "_catboost.pyx", line 3269, in _catboost._set_data
File "_catboost.pyx", line 3222, in _catboost._set_data_from_generic_matrix
File "_catboost.pyx", line 2238, in _catboost.get_cat_factor_bytes_representation
File "_catboost.pyx", line 1857, in _catboost.get_id_object_bytes_string_representation
TypeError: must be real number, not NoneType

catboost version: tried 1.0.6, 1.2.0 and 1.2.3
Operating System: Linux
CPU: yes
GPU: no

[Evgueni-Petrov-aka-espetrov]

hi!
could you please post an end-to-end reproducer?
the thing below works fine...

from catboost import CatBoostClassifier, Pool

train_data = [[0, 3],
              [4, 1],
              [8, 1],
              [9, 1]]
train_labels = [0, 0, 1, 1]

model = CatBoostClassifier(
        iterations=1000,
)
model.fit(
        train_data,
        train_labels,
        verbose=False,
        cat_features=[0]
)

pred = model.predict_proba(train_data)

pool = Pool(data=train_data, label=None, cat_features=model.get_cat_feature_indices())