corella

Correlation plots on the terminal, for quick and simple insights. Powered by pandas.

Dataset source: FIFA World Cup 2018 Match Stats

Quickstart

Requires Python3. Currently, only CSV files are supported.
The dataset parsing and correlation calculation is all done via pandas.

pip install corella

Usage

Command Line

corella accepts CSV data either through STDIN or as an input file.

cat file.csv | corella or corella --input file.csv

Supported CLI options

--pos-color

Set the color to use for positive correlation. Defaults to light_red
Supported colors include black, red, green, yellow, blue, magenta, cyan, white and light_gray, dark_gray, light_red, light_green, light_yellow, light_blue, light_magenta, light_cyan.
Full list of all 256 supported colors can be found at the colored project page.

--neg-color

Set the color to use for negative correlation. Defaults to light_blue

--input

Optional input CSV file. If specified, STDIN is ignored.

--delimiter

The delimiter for the CSV file. Defaults to ,.

--method

The method to use for calculating the correlation coefficient. Defaults to pearson.
https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.corr.html

--padding

The character padding to use on the left of the plot. Defaults to 20. If the length of a column name exceeds the specified padding, it is ignored.

--no-header

A flag to specify if the provided input does not have a header. Defaults to False. If specified, column numbers are used as names instead.

Usage within Python

from corella import Corella
C = Corella()
C.draw(pandas_df)

Sample datasets

Two datasets are provided within ./datasets/

1. The Iris dataset
2. FIFA World Cup 2018 Match Stats from Kaggle.

Try corella --input datasets/iris.csv --pos-color light_magenta --neg-color blue --padding 30

License

MIT

Thanks

colored