This module provides a handful of functions to simplify the typical data processing operations and simplifying data verification procedures.

Dependencies

• numpy 1.17.1
• pandas 0.25.1

Installation Guide

• pip install helper-funcs

Usage

Import class "HF" from module "helper_funcs":

• from helper_funcs import HF

And then call any of the methods described below.

Methods

• df_preview(df, n_samples)

  Description

  Creates a nice summary table of your DataFrame.

  Parameters

  • df: pandas.DataFrame

    The DataFrame you want to create a preview for.

  • n_samples: int, optional (default = 2)

    Number of unique values from each column to be displayed.

  Returns

  • pandas.DataFrame containing the summary information about the passed DataFrame.

• rename_col(df, old_name, new_name)

  Description

  Renames the specified column.

  Parameters

  • df: pandas.DataFrame

    The DataFrame you want to create a preview for.

  • old_name: str

    Name of existing df column to be renamed.

  • new_name: str

    Name which will replace the old_name column name.

  Returns

  • pandas.DataFrame with the renamed column.

• columns_mismatch(col_1, col_2)

  Description

  Extracts values that are present in col_1, but not in col_2.

  Parameters

  • col_1: pandas.Series

    The Series you want to subtract values from.

  • col_2: pandas.Series

    The Series which is subtracted from col_1.

  Note: The word "subtract" is used not in arithmetical sense, but in a set difference sense.

  Returns

  • Set with values which col_1 contains and col_2 does not contain.

• df_difference(df_1, df_2)

  Description

  Extracts rows that are present in df_1, but not in df_2.

  Note: df_1 and df_2 can have different column names, but number of columns should match.

  Parameters

  • df_1: pandas.DataFrame

    The DataFrame you want to subtract values from.

  • df_2: pandas.DataFrame

    The DataFrame which is subtracted from df_1.

  Note: The word "subtract" is used not in arithmetical sense, but in a set difference sense.

  Returns

  • pandas.DataFrame with rows which df_1 contains and df_2 does not contain.

• verify_dates_integity(df, date_col)

  Description

  Checks whether there are any missing dates between earliest and latest dates from df[date_col]

  Parameters

  • df: pandas.DataFrame

    The DataFrame which after selecting values from date_col will be verified for integrity

  • date_col: str

    Name of df column that will be verified for integrity

• duplicate(df, how, n_times)

  Description

  Extends the specified DataFrame by repeating its rows.

  Parameters

  • df: pandas.DataFrame

    The DataFrame which rows you want to repeat

  • how: str

    Strategy for repeating. Should be either 'whole' (then [1,2] -> [1,2,1,2]) or 'element_wise' (then [1,2] -> [1,1,2,2])

  • n_times: int

    Number of repetitions of each row

  Returns

  • Extended pandas.DataFrame with repeated rows

• groupby_to_list(df, by_cols, col_to_list)

  Description

  Extracts values of col_to_list column that correspond to the same values in by_cols column(s) and put them to list.

  Parameters

  • df: pandas.DataFrame

    The DataFrame which you want to use

  • by_cols: list of str

    Column names that will be used as keys in df

  • col_to_list: str

    Column name which values will be put to lists

  Returns

  • pandas.DataFrame with columns [by_cols, col_to_list] so that all the values in col_to_list column are lists.

• chunkenize(data_to_split, num_chunks, df_indices, copy)

  Description

  Splits the data_to_split into list with num_chunks chunks. Can be helpful when preparing data for parallel processing.

  Parameters

  • data_to_split: pandas.DataFrame or list

    The DataFrame which you want to split in chunks

  • num_chunks: int

    Number of chunks that your data will be split in

  • df_indices: list of str, optional (default = [])

    This can be used when data_to_split is pandas.DataFrame. These column will be used as DataFrame index before splitting and will be reset afterwards.

  • copy: bool, optional (default = True)

    Determines whether you want to perform splitting on a copy of data_to_split.

  Returns

  • List of num_chunks chunks that have same type as data_to_split.

• filter_df(df, col_name, l_bound, r_bound, inclusive)

  Description

  Filters the df DataFrame col_name column so that it contains only records that corresponds to df[col_name] values in the range between l_bound and r_bound.

  Parameters

  • df: pandas.DataFrame

    The DataFrame which column col_name you want to filter

  • col_name: str

    Column name from df which values you want to filter df on

  • l_bound: same type as values of df[col_name]

    Left bound of the filtered values range. Can be omitted if r_bound is specified

  • r_bound: same type as values of df[col_name]

    Right bound of the filtered values range. Can be omitted if l_bound is specified

  • inclusive: bool, optional (default = True)

    Determines whether you want range to be inclusive (True) or exclusive (False)

  Returns

  • Filtered pandas.DataFrame

• prepare_str_cols(df, make_uppercase)

  Description

  Strips leading and trailing spaces in str columns of df and makes those values to either upper-case or lower-case.

  Parameters

  • df: pandas.DataFrame

    The DataFrame you want to prepare str columns for.

  • make_uppercase: bool

    Determines whether you want str values to be upper-cased or lower-cased.

  Returns

  • pandas.DataFrame where all strings are either upper-cased or lower-cased with all leading and trailing spaces removed.