Topsis-Dhruv-102303785 0.2

TOPSIS implementation using Python

Package Description

This package is built as a submission to Assignment 1

Submitted by: Dhruv Sethi (102303785)

Definition:
TOPSIS, acronym for Technique for Order Preference by Similarity to Ideal Solution, is a multi-criteria decision-making (MCDM) method used to rank alternatives by selecting the option that is closest to the ideal solution and farthest from the worst (negative ideal) solution.

Use this package as a WebApp here.

Installation

Use the package manager pip to install Topsis-Dhruv-102303785

pip install Topsis-Dhruv-102303785

⚙️ Usage

topsis data.csv 1,1,1,1,1 +,+,-,+,+ result.csv

where,

data.csv is your input file

Weights to be assigned to each feature

Impacts to optimize each feature (+: Maximize, -: Minimize)

Output file path to save the results

🪴 Example

Input Dataset (data.csv)

Fund Name	P1	P2	P3	P4	P5
M1	0.72	0.54	4.1	43.2	12.05
M2	0.69	0.51	3.6	59.8	15.92
M3	0.67	0.44	5.0	62.1	17.10
M4	0.78	0.61	3.9	41.4	11.80
M5	0.83	0.68	5.4	39.6	11.35
M6	0.80	0.65	5.6	55.9	15.88
M7	0.86	0.74	6.5	52.3	14.70
M8	0.92	0.85	5.7	66.2	18.30

Weights: 1,1,1,1,1
Impacts: +,+,-,+,+

---

Output after applying TOPSIS (result.csv)

Fund Name	P1	P2	P3	P4	P5	Topsis Score	Rank
M1	0.72	0.54	4.1	43.2	12.05	0.3834	7
M2	0.69	0.51	3.6	59.8	15.92	0.5435	2
M3	0.67	0.44	5.0	62.1	17.10	0.4540	5
M4	0.78	0.61	3.9	41.4	11.80	0.4385	6
M5	0.83	0.68	5.4	39.6	11.35	0.3678	8
M6	0.80	0.65	5.6	55.9	15.88	0.5044	3
M7	0.86	0.74	6.5	52.3	14.70	0.4702	4
M8	0.92	0.85	5.7	66.2	18.30	0.7073	1

Notes

The package handles the following:

• Correct number of parameters (input file, weights, impacts, output file)
• Appropriate error messages for invalid inputs
• Handling of File Not Found exception
• Input file must contain three or more columns
• From second to last columns must contain numeric values only
• Number of weights, impacts and criteria columns must be the same
• Impacts must be either + or -
• Weights and impacts must be separated by commas

License

MIT License