CODEWITHSUNDEEP
pythonpandasnumpyscikit-learndata-science
kwekie
kwekie

Reputation: 35

Error when imputing using .fit() from SimpleImputer

I am working my way through the resource Python for Data Science For Dummies. I am currently learning about imputing missing data values using pandas. Below is my code:

import pandas as pd
import numpy as np
from sklearn.impute import SimpleImputer

imp = SimpleImputer(missing_values='NaN',
              strategy='mean')
# creates imputer to replace missing values.
# missing_values parameter defines what we are looking out for to impute.
# strategy parameter implies with what value you want to replace the missing value.
# strategy can be either: mean, median, most_frequent

imp.fit([[1, 2, 3, 4, 5, 6, 7]])
'''
Before imputing, we must provide stats for the imputer to use by calling fit(). 
'''

s = [[1, 2, 3, np.NaN, 5, 6, None]]


print(imp.transform(s))
x = pd.Series(imp.transform(s).tolist()[0])  # .transform() fills in the missing values in s
# we want to display the result as a series. 
# from the imputer we want to transform our imputer output to a list using .tolist()
# after that we want to transform the list into a series by enclosing it within .Series()
print(x)

However, when I run the code, it returns an error at the line with imp.fit():

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-38-3b624663cf89> in <module>
     15 # strategy can be either: mean, median, most_frequent
     16 
---> 17 imp.fit([[1, 2, 3, 4, 5, 6, 7]])
     18 '''
     19 Before imputing, we must provide stats for the imputer to use by calling fit().

/Applications/anaconda3/lib/python3.7/site-packages/sklearn/impute/_base.py in fit(self, X, y)
    266         self : SimpleImputer
    267         """
--> 268         X = self._validate_input(X)
    269         super()._fit_indicator(X)
    270 

/Applications/anaconda3/lib/python3.7/site-packages/sklearn/impute/_base.py in _validate_input(self, X)
    242                 raise ve
    243 
--> 244         _check_inputs_dtype(X, self.missing_values)
    245         if X.dtype.kind not in ("i", "u", "f", "O"):
    246             raise ValueError("SimpleImputer does not support data with dtype "

/Applications/anaconda3/lib/python3.7/site-packages/sklearn/impute/_base.py in _check_inputs_dtype(X, missing_values)
     26                          " both numerical. Got X.dtype={} and "
     27                          " type(missing_values)={}."
---> 28                          .format(X.dtype, type(missing_values)))
     29 
     30 

ValueError: 'X' and 'missing_values' types are expected to be both numerical. Got X.dtype=float64 and  type(missing_values)=<class 'str'>.

Any help on the matter is greatly appreciated!

Also, wherever you are I hope that you are coping well with the COVID-19 situation!

Upvotes: 1

Views: 1598

Answers (1)

kederrac
kederrac

Reputation: 17322

your parameter missing_values has a string as a value 'NaN' you can use:

missing_values = np.nan

Upvotes: 1

Related Questions