102303812-topsis

A Python package to implement TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) for multi-criteria decision analysis.

Author: Harsh Tanwar
Roll No: 102303812
Email: htanwar_be23@thapar.edu

Repository: Assignment1-Topsis

---

What is TOPSIS?

TOPSIS ranks alternatives by comparing each to an ideal positive and ideal negative solution. Alternatives closer to the ideal positive and farther from the ideal negative get higher scores and better ranks. The method uses vector normalization, weighted criteria, and Euclidean distances to compute a performance score for each alternative.

---

Requirements

• Python 3.6+
• pandas – data handling
• numpy – numerical operations

Install dependencies:

pip install pandas numpy

---

Installation

From PyPI (if published)

pip install 102303812-topsis

From source (clone or download the repo)

git clone https://github.com/htan11/course-materials.git
cd course-materials/UCS654/Assignment1-Topsis
pip install .

Or, in the folder containing setup.py:

pip install .

---

Usage

Command line

topsis <InputDataFile> <Weights> <Impacts> <OutputResultFileName>

Example (with 5 criteria):

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

As a Python module

(Because the package name contains a hyphen, use importlib to import.)

import importlib
topsis = getattr(importlib.import_module("102303812-topsis"), "topsis")

topsis("data.csv", "1,1,1,1,1", "+,+,-,+,-", "result.csv")

---

Parameters

Parameter	Description
InputDataFile	Path to a CSV file: first column = alternative names, remaining columns = numeric criterion values. Must have a header row.
Weights	Comma-separated weights for each criterion (e.g. "1,1,1,2,1"). Number of values must match number of criteria.
Impacts	Comma-separated impacts: + for beneficial (higher is better), - for non-beneficial (lower is better). Same count as criteria.
OutputResultFileName	Path for the output CSV. Contains original columns plus Topsis Score and Rank.

• Beneficial (+): e.g. Quality, Customer Rating — higher value is better.
• Non-beneficial (-): e.g. Price, Delivery Time — lower value is better.

---

Input file format

CSV with a header; first column = alternative names, rest = numeric only.

Example (data.csv):

Product	Quality	Price	Features	Customer Rating	Delivery Time
A1	8.5	450	7.2	4.3	3
A2	7.8	380	8.1	4.5	5
...	...	...	...	...	...

• Column 1: labels (not used in math).
• Columns 2–6: criteria. Weights and impacts must be given for each of these in order.

---

Example with this dataset

For 5 criteria (Quality, Price, Features, Customer Rating, Delivery Time):

• Weights: "1,1,1,1,1" (equal weight).
• Impacts: "+,+,-,+,-" (Quality +, Price +, Features -, Customer Rating +, Delivery Time -).

Run:

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

Output CSV will have all input columns plus:

• Topsis Score – higher is better.
• Rank – 1 = best alternative.

---

Output

The result file contains:

• All original columns.
• Topsis Score – performance score in [0, 1].
• Rank – integer rank (1 = best).

Example message: Result file 'result.csv' created successfully.

---

Project structure

assignment1_DS-main/
├── README.md
├── setup.py
├── data.csv              # Sample input
├── 102303812-topsis/
│   ├── __init__.py
│   └── topsis.py         # TOPSIS implementation
└── LICENSE

---

License

MIT License. See LICENSE for details.