machine learning - Using sklearn Base Estimator and TransformerMixin

Question

I am working on a loan prediction problem, the dataset includes categorical and numerical features. I am writing a custom transformer that will process both numerical and categorical attributes. Funny enough I am getting a syntax error and I can't see the problem. It's my first time writing custom transformers so practical tips and correction will be appreciated

from sklearn.base import BaseEstimator, TransformerMixin
class data_cleaner(BaseEstimator, TransformerMixin):
    def __init__(self, cat_data, num_data):
        self.cat_data = cat_data
        self.num_data = num_data
        
    def process_data(self, cat_data, num_data):
        cat_data = []
        num_data = []
        for i,c in enumerate(X.dtypes):
            if c == object:
                cat_data.append(X.iloc[:, i])
            else :
                num_data.append(X.iloc[:, i]
        cat_data = pd.DataFrame(cat_data).transpose()
        num_data = pd.DataFrame(num_data).transpose()
        #filling the missing categorical data
        cat_data = cat_data.apply(lambda value:value.fillna(value.value_counts().index[0]))
        #mapping the data on the loan_status
        target = {'Y': '1', 'N': '0'}
        loan_status = cat_data['Loan_Status']
        #dropping the loan_status from cat_data
        cat_data.drop('Loan_Status', inplace=True, axis=1)
        #mapping loan_status wrt target
        loan_status = loan_status.map(target)
        #changing the cat_data into numerical values
        from sklearn.preprocessing import LabelEncoder
        encoder = LabelEncoder()
        for value in cat_data:
            cat_data[value] = encoder.fit_transform(cat_data[value])
         # Numerical data
        # Fill every missing value with their previous value in the same column.

        num_data.fillna(method='bfill', inplace=True)
        #joining the cat_data and num_data
    
    def transform(self, X, y=None):        
        X = pd.concat([cat_data, num_data, loan_status], axis=1)
                                                    
    def fit(self, *_):
        return self

question from:https://stackoverflow.com/questions/65874152/using-sklearn-base-estimator-and-transformermixin

Answer 1

Waitting for answers