Topsis-Taranpreet-102017050 0.4

Topsis implementation

Topsis_Taranpreet_102017050

TOPSIS (technique for order performance by similarity to ideal solution) is a useful technique in dealing with multi-attribute or multi-criteria decision making (MADM/MCDM) problems in the real world.

Installation

Install the Package using the command -

$ pip install Topsis_Taranpreet_102017050

Usage

python -m Topsis_Taranpreet_102017050.Topsis <InputDataFile> <Weights> <Impacts> <ResultFileName>

Example

Fund Name	P1	P2	P3	P4	P5
M1	0.94	0.88	6.1	40.1	12.01
M2	0.85	0.72	5	52.9	14.87
M3	0.84	0.71	4.5	43.2	12.31
M4	0.73	0.53	6.7	43.8	12.94
M5	0.88	0.77	6.5	31.7	9.96
M6	0.6	0.36	4.4	31.2	9.14
M7	0.6	0.36	4.4	48	13.34
M8	0.92	0.85	5.5	55.2	15.62

Input Method: The Input data file is data.csv, output file is result.csv and for the Weights : [2, 2, 3, 3, 4] & Impacts : [-, +, -, +, -], run the following command:

python -m Topsis_Taranpreet_102017050.Topsis data.csv "2,2,3,3,4" "-,+,-,+,-" result.csv

Output generated: The output file, result.csv will be as follows:

Fund Name	P1	P2	P3	P4	P5	Performance Score	Rank
M1	0.94	0.88	6.1	40.1	12.01	0.522346027	3
M2	0.85	0.72	5	52.9	14.87	0.508871391	4
M3	0.84	0.71	4.5	43.2	12.31	0.580231695	1
M4	0.73	0.53	6.7	43.8	12.94	0.390517293	8
M5	0.88	0.77	6.5	31.7	9.96	0.503787007	5
M6	0.6	0.36	4.4	31.2	9.14	0.531336089	2
M7	0.6	0.36	4.4	48	13.34	0.493355187	7
M8	0.92	0.85	5.5	55.2	15.62	0.50091679	6

Points to note:

• To remove the indices and headers, the library implicitly removes the first column and row respectively. Kindly, make sure the csv follows the format as shown in sample.csv.
• The csv should not contain categorical values.
• The csv should have atleast more than 3 columns.
• The number of Impacts an Weights should be equal to the number of feature columns.
• For maximizing a column, the impact is shown by "+" and for minimizing, "-".
• The weights should be positive and numerical.
• Separate the weights and columns by comma (,).
• Please follow the format to run the program as given in the sample command.

License

MIT

Thank You! Keep using and sharing feedbacks!