Why does tuple come out from LGBM?

Question

from lightgbm import LGBMClassifier
model = LGBMClassifier()
model.fit(X_train, y_train)

This is my code. I got this error in model.fit():

  AttributeError                            Traceback (most recent call last)
    <ipython-input-76-6b35b759d36e> in <module>
          1 from lightgbm import LGBMClassifier
          2 model = LGBMClassifier()
    ----> 3 model.fit(X_train, y_train)
          4 y_pred = model.predict_proba(X_valid)

/usr/local/lib/python3.6/dist-packages/lightgbm/basic.py in c_str(string)
    110 def c_str(string):
    111     """Convert a Python string to C string."""
--> 112     return ctypes.c_char_p(string.encode('utf-8'))
    113 
    114 
    
    AttributeError: 'tuple' object has no attribute 'encode'

X_train is a DataFrame and has (2500,56) shape that all the dtypes are float64.

y_train is a Series and has (2500,) shape that the dtype is int64.

It worked well in other models like Xgb and RandomForest.

You'd need to provide the error traceback, and at least some sense of what `X_train, y_train` look like. — Ben Reiniger, Feb 10 '21 at 21:54

score 1 · Answer 1 · answered Feb 11 '21 at 10:42

1

I found what I was missing.

X_train was a DataFrame that has multi-index.

X_train.columns = X_train.columns.get_level_values(0)

This code solved the problem.

answered Feb 11 '21 at 10:42

OZOOOH

31
5

Why does tuple come out from LGBM?

1 Answers1