5 examples of 'pandas transform' in Python

51
def transform(self, X, columns=None):
52
    self._check_input(X)
53
    x_type = type(X)
54
55
    if x_type == MultiSeries:
56
        return self._transform_series(X)
57
58
    # If X is not MultiSeries, then it's MultiDataFrame
59
    # Because of self._check_input function
60
    return self._transform_data_frame(X)

140
def transform(self, X, columns_mapping=None):
141
    '''
142
    Transform X with fitted dictionary self.transformations.
143
    :param columns_mapping: {old_col: new_col} mapping between columns in fit data set and current X
144
    :return:
145
    '''
146
    if columns_mapping is None:
147
        columns_mapping = {}
148
149
    transformers_df = X.copy()
150
151
    for col_name, transformations in self.transformations.items():
152
        for t in transformations:
153
            new_col_name = columns_mapping.get(col_name, col_name)
154
            transformed_column = t.transform(X[new_col_name])
155
156
            if type(transformed_column) == XSeries:
157
                transformers_df.rename(columns={
158
                    new_col_name: transformed_column.name
159
                }, inplace=True)
160
                transformers_df[transformed_column.name] = transformed_column
161
            else:
162
                transformers_df.drop(new_col_name, inplace=True, axis=1)
163
164
                transformers_df = XDataFrame.concat_dataframes(
165
                    [transformers_df, transformed_column]
166
                )
167
168
    return transformers_df

44
def transform(self, data):
45
    """User implements internal transform function which operates on a single data element from a data_object
46
    
47
    Args:
48
        data (object): input data
49
50
    Returns:
51
        data, transformed
52
53
    """
54
    if isinstance(data, pd.DataFrame):
55
        if self.columns:
56
            scaled_features_df = data.copy()
57
58
            # transform select columns
59
            scaled_features = self.preprocessor.transform(data[self.columns].values)
60
            for idx, colname in enumerate(self.columns):
61
                scaled_features_df[colname] = scaled_features[:,idx]
62
63
        else:
64
            scaled_features = self.preprocessor.transform(data.values)
65
            try:
66
                scaled_features_df = pd.DataFrame(scaled_features, index=data.index, columns=data.columns)
67
            except ValueError:
68
                logging.info(f'{self.preprocessor.__class__.__name__} instance changed the number of columns. Returning raw values')
69
                return pd.DataFrame(scaled_features)
70
        return scaled_features_df
71
72
    return self.preprocessor.transform(data)

250
def _transform(self, dataset):
251
    self._transfer_params_to_java()
252
    return DataFrame(self._java_obj.transform(dataset._jdf), dataset.sql_ctx)

7
def dataframe_transform_parallel(
8
        df, transformer
9
                   ):
10
    cpu_count = multiprocessing.cpu_count()
11
    workers_count = int(round(cpu_count))
12
    logger.log(15, 'Dataframe_transform_parallel running pool with '+str(workers_count)+' workers')
13
    df_chunks = np.array_split(df, workers_count)
14
    df_list = execute_multiprocessing(workers_count=workers_count, transformer=transformer, chunks=df_chunks)
15
    df_combined = pd.concat(df_list, axis=0, ignore_index=True)
16
    return df_combined