topsis-arnav-102203979 1.4

Python implementation of the TOPSIS method for multi-criteria decision analysis

TOPSIS Algorithm - Ranking Made Simple

Welcome to this Python implementation of TOPSIS (Technique for Order Preference by Similarity to Ideal Solution), a method used to rank alternatives based on a set of decision-making criteria. This project aims to simplify complex decision-making processes by using mathematical formulas to calculate the best options.

How Does It Work?

Imagine you are comparing different models of smartphones based on various features like price, storage, camera quality, and design. Instead of relying on your gut feeling, TOPSIS normalizes the data, assigns weights to each feature (e.g., you care more about camera quality than looks), and then calculates the distance between each model and an ideal solution. The closer a model is to the "best-case scenario," the better its rank.

What You Will Need

To get started, you will need:

• Python 3.7+ (preferably)
• Pandas for data handling
• NumPy for numerical calculations

To install the dependencies, just run:

pip install pandas numpy

How to Run the Script

Command-Line Execution

Here is the command you will use to run the script:

topsis <inputFileName> <Weights> <Impacts> <resultFileName>

Where:

• <inputFileName> is your CSV file that holds the data. It should have one column for alternatives and the rest for decision-making criteria (e.g., price, storage, camera).
• <Weights> is a comma-separated string that specifies the importance (weight) of each criterion.
• <Impacts> is another comma-separated string, indicating whether a criterion is beneficial (+) or non-beneficial (-).
• <resultFileName> is where you want to save the output, containing the scores and ranks.

Example

Let us say we have a CSV file named input_data.csv and we want to weigh the features as 20%, 30%, 30%, and 20%, respectively, with the first three being beneficial and the last one (Looks) being non-beneficial.

Run:

topsis input_data.csv "0.2,0.3,0.3,0.2" "+,+,+,-" result.csv

This command will process the data, calculate the scores, and save the results in a file called result.csv.

Input File Structure

The CSV file should look like this:

Model	Price	Storage	Camera	Looks
M1	250	16	12	5
M2	200	16	8	3
M3	300	32	16	4
M4	275	32	8	4
M5	225	16	16	2

Here, each row represents a different alternative (e.g., smartphone model) and each column represents a decision-making criterion.

Output Format

After running the script, the output will look like this:

Model	Price	Storage	Camera	Looks	Score	Rank
M1	250	16	12	5	0.4459	4
M2	200	16	8	3	0.3700	5
M3	300	32	16	4	0.6299	1
M4	275	32	8	4	0.4936	2
M5	225	16	16	2	0.4758	3

The Score is a measure of how close each alternative is to the ideal solution, and the Rank shows the order of preference, with 1 being the best choice.

Handling Errors

The script takes care of common mistakes:

• If your input file does not exist, it will raise a FileNotFoundError.
• If you mismatch the number of weights and impacts, or if you provide incorrect impacts (e.g., using anything other than + or -), you will get a ValueError.

License

This project is licensed under the MIT License. Feel free to use, modify, or contribute!