AttributeError: 'numpy.ndarray' object has no attribute 'columns'

This looks amazing but I can’t run it on my dataset. I get the following error:

runfile('E:/Machine Learning Projects/ML UFC/genetic_algorithm_cluster_UFC_database_TPOT.py', wdir='E:/Machine Learning Projects/ML UFC')
Traceback (most recent call last):

  File "E:\Machine Learning Projects\ML UFC\genetic_algorithm_cluster_UFC_database_TPOT.py", line 533, in <module>
    ExplainerDashboard(ClassifierExplainer(RandomForestClassifier().fit(X_train, y_train), X_test, y_test)).run()

  File "C:\Users\apavl\anaconda3\envs\neuro\lib\site-packages\explainerdashboard\explainers.py", line 1774, in __init__
    super().__init__(model, X, y, permutation_metric,

  File "C:\Users\apavl\anaconda3\envs\neuro\lib\site-packages\explainerdashboard\explainers.py", line 149, in __init__
    self.onehot_cols, self.onehot_dict = parse_cats(self.X, cats)

  File "C:\Users\apavl\anaconda3\envs\neuro\lib\site-packages\explainerdashboard\explainer_methods.py", line 103, in parse_cats
    all_cols = X.columns

AttributeError: 'numpy.ndarray' object has no attribute 'columns'

Any idea why that is?

[oegedijk]

Hi @apavlo89, I agree that it would be cool to support sklearn pipelines and I’ve tried to get them to work, but the problem is with how they are internally structured: the stages of the pipeline forward numpy arrays instead of dataframes, and it doesn’t track what the newly generated features should be called.

There is a proposed fix for this with here: https://scikit-learn-enhancement-proposals.readthedocs.io/en/latest/slep007/proposal.html

Given that the dashboard needs to know the names of the features in order to give meaningful importance and dependence plots, that means that until that proposal is implemented it will only accept pipelines that do not add or remove any columns.

[aakash1998]

def vif_score(x): sc = StandardScaler() arr = sc.fit_transform(x) return pd.DataFrame([[x.columns[i] , variance_inflation_factor(arr , i)] for i in range(arr.shape[1])] , columns = [‘Features’ , ‘VIF’])

I am trying to run this function … earlier it worked fine , but don’t know what’s the issue , it shows the error “‘numpy.ndarray’ object has no attribute ‘columns’”.