NameError: name 'PY3' is not defined

[MSusik]

Hi!

I installed catboost on Python 3.6.3 within a Docker container (python:3.6 image). The installation was done through PyPI. The library version is 0.6.3

I can import catboost, but when I do:

model = catboost.CatBoostRegressor()

I obtain an error:

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-16-e4bd7d18c439> in <module>()
----> 1 model = catboost.CatBoostRegressor()

/usr/local/lib/python3.6/site-packages/catboost/core.py in __init__(self, iterations, learning_rate, depth, l2_leaf_reg, model_size_reg, rsm, loss_function, border_count, feature_border_type, fold_permutation_block_size, od_pval, od_wait, od_type, nan_mode, counter_calc_method, leaf_estimation_iterations, leaf_estimation_method, thread_count, random_seed, use_best_model, verbose, logging_level, metric_period, ctr_leaf_count_limit, store_all_simple_ctr, max_ctr_complexity, has_time, one_hot_max_size, random_strength, name, ignored_features, train_dir, custom_metric, eval_metric, bagging_temperature, save_snapshot, snapshot_file, fold_len_multiplier, used_ram_limit, gpu_ram_part, allow_writing_files, approx_on_full_history, boosting_type, simple_ctr, combinations_ctr, per_feature_ctr, ctr_description, task_type, device_config, devices, bootstrap_type, max_depth, n_estimators, num_boost_round, num_trees, colsample_bylevel, random_state, reg_lambda, objective, eta, max_bin, **kwargs)
   1767             objective = None
   1768 
-> 1769         if isinstance(loss_function, str) and self._is_classification_loss(loss_function):
   1770             raise CatboostError("Invalid loss_function={}: for Regressor use RMSE, MAE, Quantile, LogLinQuantile, Poisson, MAPE, R2.".format(loss_function))
   1771         params = {}

_catboost.pyx in _catboost._CatBoostBase._is_classification_loss()

_catboost.pyx in _catboost.to_binary_str()

NameError: name 'PY3' is not defined

Same happens when I specify a loss function:

model = catboost.CatBoostRegressor(loss_function='MAPE')

[aldanor]

Probably caused by #268 - try import catboost._catboost and see what happens.

[victorzhrn]

same problem here

[aldanor]

See the comment above

[victorzhrn]

Tried, still error.

---

AttributeError Traceback (most recent call last)
in ()
----> 1 import catboost._catboost

_catboost.pyx in init _catboost()

AttributeError: type object '_catboost._FloatArrayWrapper' has no attribute 'reduce_cython'

[xmoanvaf]

same problem, i tried import catboost._catboost and get message below

---

  1 from catboost import CatBoostRegressor
  2 import numpy

----> 3 import catboost._catboost

_catboost.pyx in init _catboost()

AttributeError: type object '_catboost._FloatArrayWrapper' has no attribute 'reduce_cython'

[esceptico]

try to 'pip install PY3'
worked for me

[esceptico]

wow, on my home pc my previous answer doesn't work

[sharthZ23]

Same problem, nothing above didnt work for me

[annaveronika]

Sorry, last comment was for a different issue. This one is not resolved jet.

[Yura52]

+1
That's how I faced the same problem:

features  = <some dataframe>
target = <some series>
p = catboost.Pool(features, target)

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-224-a3edbf41d908> in <module>()
----> 1 catboost.Pool(f, t)

~/miniconda3/envs/tafd/lib/python3.6/site-packages/catboost/core.py in __init__(self, data, label, cat_features, column_description, pairs, delimiter, has_header, weight, group_id, subgroup_id, pairs_weight, baseline, feature_names, thread_count)
    206                 self._read(data, column_description, pairs, delimiter, has_header, thread_count)
    207             else:
--> 208                 self._init(data, label, cat_features, pairs, weight, group_id, subgroup_id, pairs_weight, baseline, feature_names)
    209         super(Pool, self).__init__()
    210 

~/miniconda3/envs/tafd/lib/python3.6/site-packages/catboost/core.py in _init(self, data_matrix, label, cat_features, pairs, weight, group_id, subgroup_id, pairs_weight, baseline, feature_names)
    518         if feature_names is not None:
    519             self._check_feature_names(feature_names, features_count)
--> 520         self._init_pool(data_matrix, label, cat_features, pairs, weight, group_id, subgroup_id, pairs_weight, baseline, feature_names)
    521 
    522 

_catboost.pyx in _catboost._PoolBase._init_pool()

_catboost.pyx in _catboost._PoolBase._init_pool()

_catboost.pyx in _catboost._PoolBase._set_feature_names()

_catboost.pyx in _catboost.to_binary_str()

NameError: name 'PY3' is not defined

Ubuntu 16.04
All the software (python 3.6.5, numpy, six, catboost, whatever else) is of the latest versions currently available and was installed via pip 9.0.3

UPDATE:
I noticed that the error took place in the method _set_feature_names . So I replaced features with features.values and it fixed the problem.

[jrzaurin]

in my case appears for a different reason, not at _set_feature_names

clf = ctb.CatBoostClassifier(eval_metric='F1', scale_pos_weight=9.)
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-9-15bc93ece38c> in <module>()
----> 1 clf = ctb.CatBoostClassifier(eval_metric='F1', scale_pos_weight=9)

~/.virtualenvs/py3/lib/python3.6/site-packages/catboost/core.py in __init__(self, iterations, learning_rate, depth, l2_leaf_reg, model_size_reg, rsm, loss_function, border_count, feature_border_type, fold_permutation_block_size, od_pval, od_wait, od_type, nan_mode, counter_calc_method, leaf_estimation_iterations, leaf_estimation_method, thread_count, random_seed, use_best_model, verbose, logging_level, metric_period, ctr_leaf_count_limit, store_all_simple_ctr, max_ctr_complexity, has_time, classes_count, class_weights, one_hot_max_size, random_strength, name, ignored_features, train_dir, custom_loss, custom_metric, eval_metric, bagging_temperature, save_snapshot, snapshot_file, fold_len_multiplier, used_ram_limit, gpu_ram_part, allow_writing_files, approx_on_full_history, boosting_type, simple_ctr, combinations_ctr, per_feature_ctr, ctr_description, task_type, device_config, devices, bootstrap_type, max_depth, n_estimators, num_boost_round, num_trees, colsample_bylevel, random_state, reg_lambda, objective, eta, max_bin, scale_pos_weight, **kwargs)
   1548             objective = None
   1549
-> 1550         if isinstance(loss_function, str) and not self._is_classification_loss(loss_function):
   1551             raise CatboostError("Invalid loss_function='{}': for classifier use "
   1552                                 "Logloss, CrossEntropy, MultiClass, MultiClassOneVsAll, AUC, Accuracy, Precision, Recall, F1, TotalF1, MCC or custom objective object".format(loss_function))

_catboost.pyx in _catboost._CatBoostBase._is_classification_loss()

_catboost.pyx in _catboost.to_binary_str()

NameError: name 'PY3' is not defined

I am on CentOS7, Python 3.6.4 and updated versions for everything else

[aldanor]

Guys, as I’ve already replied above, this error is the consequence (of the extension not being fully imported) and not the underlying cause, see #268.

[jrzaurin]

Hey @aldanor

ok, let me check there. Sorry for the repetition

[shaz13]

Same issue here