Topsis-Tanvi-102213024 1.0.2

A Python package to perform TOPSIS (Technique for Order Preference by Similarity to Ideal Solution)

Topsis_Tanvi_102213024

Topsis_Tanvi_102213024is a Python package for implementing the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). It is a popular multi-criteria decision-making method used to rank alternatives based on their relative closeness to the ideal solution.

Installation

You can install the package using pip:

pip install Topsis_Tanvi_102213024

Usage

Input File Format

The input CSV file must follow this structure:

• The first column should contain the names of the alternatives (e.g., products, models, or options).
• Subsequent columns should contain the numerical values of the criteria for each alternative.
• The first row should provide headers for all columns.

Example

Suppose you have an input CSV file named data.csv with the following content:

Model	Price	Battery Life	Performance	Portability
m1	300	10	8	6
m2	250	8	7	9
m3	400	9	9	5

You want to apply the TOPSIS method with the following parameters:

• Weights for the criteria: Price (0.5), Battery Life (0.3), Performance (0.2)
• Impacts for the criteria: Price (-), Battery Life (+), Performance (+)

The command would be:

topsis data.csv "0.5,0.3,0.2" "-,+,+" results.csv


### Command-line Usage

After installation, you can use the `topsis` command in the terminal:

```bash
topsis <input_file> <weights> <impacts> <output_file>

• <input_file>: Path to the input CSV file.
• <weights>: Comma-separated string of weights for the criteria.
• <impacts>: Comma-separated string of '+' or '-' indicating the desirability of the criteria.
• <output_file>: Path to the output CSV file to save the results.

Example

topsis data.csv "0.5,0.3,0.2" "+,-,+" results.csv

This command will process the data.csv file using the specified weights and impacts and output the results to results.csv.

NOTE

• Ensure the input file contains at least three columns: one for alternatives and at least two for criteria.
• All criteria values should be numerical.
• The number of weights and impacts should match the number of criteria columns.
• Impacts must only include + (beneficial) or - (non-beneficial).

License

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