Topsis-Prabhpreet-102303258 0.1.0

A command line implementation of TOPSIS

Topsis-Prabhpreet-102303258

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

---

Package Structure

Topsis-Prabhpreet-102303258/
│
├── topsis_prabhpreet_102303258/
│   ├── __init__.py
│   └── topsis.py
│
├── README.md
├── requirements.txt
├── pyproject.toml
├── LICENSE
└── sample_data.csv

---

How TOPSIS Works

1. Normalize the decision matrix using vector normalization
2. Weight the normalized matrix with user-defined weights
3. Identify ideal best and ideal worst solutions for each criterion
4. Calculate Euclidean distances from ideal solutions
5. Compute relative closeness score (TOPSIS Score)
6. Rank alternatives based on their scores (higher = better)

---

Installation

From PyPI (Recommended)

pip install Topsis-Prabhpreet-102303258

From Source

git clone https://github.com/PrabhpreetSingh/Topsis-Prabhpreet-102303258.git
cd Topsis-Prabhpreet-102303258
pip install .

---

Command-Line Usage

Syntax

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

Parameters

Parameter	Description
InputDataFile	Path to input CSV file containing the decision matrix
Weights	Comma-separated numeric weights for each criterion
Impacts	Comma-separated impacts: + (beneficial) or - (non-beneficial)
OutputResultFileName	Path for the output CSV file with results

Example

topsis sample_data.csv "1,1,1,2" "+,+,-,+" result.csv

This command:

• Reads sample_data.csv as input
• Applies weights [1, 1, 1, 2] to four criteria
• Uses impacts [+, +, -, +] (maximize 1st, 2nd, 4th; minimize 3rd)
• Saves results to result.csv

---

Input File Format

The input CSV file must follow this structure:

Object,Criterion1,Criterion2,Criterion3,Criterion4
M1,0.79,0.62,1.25,38.5
M2,0.66,0.44,2.89,63.7
M3,0.56,0.31,1.57,42.1
M4,0.82,0.67,2.68,74.2
M5,0.75,0.56,1.83,51.9

Requirements

• First column: Object/alternative names (can be text)
• Columns 2 onwards: Numeric criteria values only
• Minimum: 3 columns total (1 name + 2 criteria)

---

Output Description

The output CSV contains all original columns plus:

Column	Description
Topsis Score	Relative closeness coefficient (0 to 1, higher is better)
Rank	Ranking position (1 = best)

Example Output

Model,Corr,R2,RMSE,Accuracy,Topsis Score,Rank
M1,0.79,0.62,1.25,38.5,0.534,3
M2,0.66,0.44,2.89,63.7,0.308,5
M3,0.56,0.31,1.57,42.1,0.373,4
M4,0.82,0.67,2.68,74.2,0.695,1
M5,0.75,0.56,1.83,51.9,0.535,2

---

Error Handling

The program validates all inputs and provides clear error messages:

Error	Cause
Incorrect number of parameters	Not exactly 4 command-line arguments provided
Input file not found	Specified input file does not exist
Must contain at least three columns	Input file has fewer than 3 columns
Must contain numeric values only	Non-numeric values found in criteria columns
Number of weights must equal criteria	Mismatch between weights and number of criteria
Number of impacts must equal criteria	Mismatch between impacts and number of criteria
Impact must be '+' or '-'	Invalid impact character used

Example Error Messages

# Wrong number of arguments
$ topsis data.csv "1,1,1"
Error: Incorrect number of parameters.
Usage: topsis <InputDataFile> <Weights> <Impacts> <OutputResultFileName>

# File not found
$ topsis missing.csv "1,1,1,1" "+,+,-,+" out.csv
Error: Input file 'missing.csv' not found.

# Invalid impact
$ topsis data.csv "1,1,1,1" "+,+,x,+" out.csv
Error: Impact at position 3 must be '+' or '-', got 'x'.

---

Complete Example

Step 1: Create Input File

Create data.csv:

Model,Corr,R2,RMSE,Accuracy
M1,0.79,0.62,1.25,38.5
M2,0.66,0.44,2.89,63.7
M3,0.56,0.31,1.57,42.1
M4,0.82,0.67,2.68,74.2
M5,0.75,0.56,1.83,51.9

Step 2: Run TOPSIS

topsis data.csv "1,1,1,2" "+,+,-,+" result.csv

Explanation:

• Corr (weight=1, impact=+): Higher correlation is better
• R2 (weight=1, impact=+): Higher R² is better
• RMSE (weight=1, impact=-): Lower RMSE is better
• Accuracy (weight=2, impact=+): Higher accuracy is better (double weight)

Step 3: View Results

cat result.csv

Output shows M4 ranked #1 with highest TOPSIS score.

---

Dependencies

• Python >= 3.8
• pandas >= 1.3.0
• numpy >= 1.21.0

---

License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Author

Prabhpreet Singh
Roll Number: 102303258
Thapar Institute of Engineering and Technology
Email: [e-mail]

---

Links

• PyPI: https://pypi.org/project/Topsis-Prabhpreet-102303258/
• GitHub: https://github.com/PrabhpreetSingh/Topsis-Prabhpreet-102303258