GitLocker: The Coding Marketplace

Description:

datacleaning 1.0.1

DATA CLEANING
## Description
In any Machine Learning process, Data Preprocessing is the primary step wherein the raw/unclean data are transformed
into cleaned data, So that in the later stage, machine learning algorithms can be applied. This python paackage make the
data preprocessing very easy in just 2 lines of code. All you have to do is just input a raw data(CSV file), this library
will clean your data and return you the cleaned dataframe on which further you can apply feature engineering,
feature selection and modeling.

What this does?

Cleans special character
Removes duplicates
Fixes abnormality in column names
Imputes the data (categorical & numerical)

Data Cleaning
Data-cleaning is a python package for data preprocessing. This cleans the CSV file and returns the cleaned data frame.
It does the work of imputation, removing duplicates, replacing special characters, and many more.
How to use:
Step 1:
Install the libaray
pip install data-cleaning

Step 2:
Import the library, and specify the path of the csv file.
from datacleaning import DataCleaning

dp = DataCleaning(file_uploaded='filename.csv')
cleaned_df = dp.start_cleaning()

There are some optional parameters that you can specify as listed below,
Usage:
from datacleaning import DataCleaning

DataCleaning(file_uploaded='filename.csv', separator=",", row_threshold=None, col_threshold=None,
special_character=None, action=None, ignore_columns=None, imputation_type="RDF")

Parameters

Parameter
Default Value
Limit
Example

file_uploaded
none
Provide a CSV file.
filename.csv

separator
,
Separator used in csv file
;

row_threshold
none
0 to 100
80

col_threshold
none
0 to 100
80

special_character
Check the list below
Sspecify the character that is not listed in default_list (see below)
[ '$' , '?' ]

action
none
add or remove
add

ignore_columns
none
Provide list of column names to ignoring the special characters operation.
[ 'column1', 'column2' ]

imputation_type
RDF
Select your preferred imputation RDF, KNN, mean, median, most_frequent, constant .
KNN

Examples of using parameters
- Appending extra special characters to the existing default_list
The DEFAULT SPECIAL CHARACTERS included in the package are shown below,
default_list = ["!", '"', "#", "%", "&", "'", "(", ")",
"*", "+", ",", "-", ".", "/", ":", ";", "<",
"=", ">", "?", "@", "[", "\\", "]", "^", "_",
"`", "{", "|", "}", "~", "â€“", "//", "%*", ":/", ".;", "Ã˜", "Â§",'$',"Â£"]

How to remove a special character, say for example if you want to remove "?" and "%".
Note:- Do not forget to give action = 'remove'
from datacleaning import DataCleaning

dp = DataCleaning(file_uploaded='filename.csv', special_character =['?', '%'], action='remove')
cleaned_df = dp.start_cleaning()

How to add a special character, say for example if you want to add "Ã©" that is not in the default_list given above.
Note:- Do not forget to give action = 'add'
from datacleaning import DataCleaning

dp = DataCleaning(file_uploaded='filename.csv', special_character =['Ã©'], action='add')
cleaned_df = dp.start_cleaning()

- Ignoring a particular columns and adding a special character
Say for example, column named "timestamp" and "date" needs to be removed and a special character needs to be added 'Ã©'
from datacleaning import DataCleaning

dp = DataCleaning(file_uploaded='filename.csv', special_character =['Ã©'],
action='add', ignore_columns=['timestamp', 'date'])
cleaned_df = dp.start_cleaning()

- Changing threshold to remove null rows/columns above this given threshold value
from datacleaning import DataCleaning

dp = DataCleaning(file_uploaded='filename.csv', row_threshold=50, col_threshold=90)
cleaned_df = dp.start_cleaning()

- Imputation methods available

RDF (RandomForest) -> (DEFAULT)
KNN (k-nearest neighbors)
mean
median
most_frequent
constant

# Example for KNN imputation.
from datacleaning import DataCleaning

dp = DataCleaning(file_uploaded='filename.csv', imputation_type='KNN')
cleaned_df = dp.start_cleaning()

>> THANK YOU <<