Topsis-Amrita-102017017 0.0.4

A command line tool to perform TOPSIS, a multi-criteria decision making technique

TOPSIS

TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) is a decision-making technique applied in order to rank potential solutions on the basis of multiple criteria.

Installation

This package requires Python v3.5+ to run. Use pip to install:

pip install topsis-amrita-102017017

OS Compatibility

It should work on any Python implementation and operating system and is compatible with Python version 3.5 and upwards.

Usage

Run topsis in the input file's directory as follows:

topsis <input_file_name> <weights> <impacts> <output_file_name>

For example,

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

Use quotation marks while including spaces in any argument:

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

• Input and output file format should be .CSV
• First column in the input file should be the object name
• Input file must have at least 2 criteria, and all criterion values should be numeric
• Weights must be numeric and comma-separated. For example, 0.25,0.25,1.0,0.25 or "0.25,0.25,1.0,0.25".
• Impacts must be comma-separated with + for criteria that are to be maximised, and - for criteria that are to be minimised. For example, +,-,+,- or "+, -, +, -"

Example

Consider input.csv:

Model	Corr	R2	RMSE	Accuracy
M1	0.79	0.62	1.25	60.89
M2	0.66	0.44	2.89	63.07
M3	0.56	0.31	1.57	62.87
M4	0.82	0.67	2.68	70.19
M5	0.75	0.56	1.3	80.39

If we run the following command:

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

we get a file named result.csv in the directory with an additional 2 columns containing the TOPSIS score and the rank of each object:

Model	Corr	R2	RMSE	Accuracy	TOPSIS Score	Rank
M1	0.79	0.62	1.25	60.89	0.7722097345612788	2.0
M2	0.66	0.44	2.89	63.07	0.22559875426413367	5.0
M3	0.56	0.31	1.57	62.87	0.43889731728018605	4.0
M4	0.82	0.67	2.68	70.19	0.5238778712729114	3.0
M5	0.75	0.56	1.3	80.39	0.8113887082429979	1.0

License

MIT