TOPSIS implementation in python for multi-criteria decision making

Class project for DATA ANALYSIS AND VISUALISATION 2020 - UCS633 Thapar University, Patiala
Navkiran Singh  

COE17
Roll number: 101703365

Update: Added windows command line support

Output is a dataframe with 3 columns

• Alternatives serial number
• Corresponding performance Score or closeness to ideal solution
• Rank

Installation

pip install topsis_navkiran

Note the name has an underscore not a hyphen. If installation gives error or package is not found after installing, install as sudo.

Recommended - test it out in a virtual environment.

Upgrade

pip install topsis_navkiran --upgrade

To use via command line

topsis_navkiran_cli data.csv 25,25,25,25 -+++

First argument after run.py is filename with .csv extension. The .csv file is assumed to have a structure similar to one provided in topsis_navkiran/data.csv

That is, the .csv file should have a header with column names and first column should only list alternatives and not attribute values.

To use in .py script

from topsis_navkiran import topsis
"""
decision_matrix is 2D numpy array, weights is a 1D array and impacts is a string of the form +-+-- 
where + indicates benefit and - indicates cost
"""
output_dataframe = topsis(decision_matrix,weights,impacts)

Debugging and Exception Handling

The program has several assert statements which raise errors with helpful description in the following cases:

• Wrong dimensions of decision matrix (not 2D), weights (not 1D)
• Length of weights and impacts don't match
• Weights or impacts don't match number of attributes
• For command line, number of arguments is less than 3 required
• File extension must be .csv

Based off on a similar package for TOPSIS in R