Creating data

Data Preparation

Let’s start by preparing the data — this is the essential first step in any multi-criteria decision analysis. Your need to ensure that the results are meaningful and useful for decision-making.

info

Raw data should be not normalized, during the method execution data will be normalized depends on method configuration

Decision Matrix ($M$)

The matrix is built as follows:

• rows : alternatives
• columns : decision criterion

double[,] matrix = 
{
    { 120, 1500 }, 
    { 2000, 1800 }
}

Weights Vector ($W$)

The weights have to meet the assumption $\sum_{i=1}^{n} w_i = 1$ where $n$ is the number of criteria (i.e. the number of columns in the decision matrix).

double[] weights = [0.5,0.5]

Criteria Decision Vector ($C$)

They will represent as vector of $\{ c_1, c2, \dots, cn \} \in [-1, 1]$, where $n$ is the number of criteria.

• 1 for benefit criteria (the higher, the better)
• -1 for cost criteria (the lower, the better)

int[] criteriaTypes = [-1,-1]

Build data

The DataBuilder provides a fluent interface for constructing the input data used across decision-making methods. It encapsulates the definition of:

• decision criteria types,
• associated weights,
• the decision matrix containing performance scores for each alternative.

var data = new DataBuilder()
    .AddWeights(weights)
    .AddDecisionCriteria(types)
    .AddDecisionMatrix(matrix)
    .Build();

The builder enforces data consistency at build time — for example, it checks that:

• the number of weights matches the number of criteria,
• each row in the matrix corresponds to a single alternative,
• each row contains scores for all criteria.

Use this builder to safely construct the shared input data required by all MCDA methods.

warning

If you don't use any of the methods, an exception will happen.