2.4. Statistical treatments and operations

This section presents the statistical computations allowed on the attributes in a TDataServer through the following five main groups of method.

All the methods have been adapted to cope with the case of vector: a check is performed, allowing the computation to be done only if the number of element in the vector is constant (at least throughout the selection if one is requested). If the constant-size criterion is not fulfilled, the considered vector is disregarded for this method. The methods detailed here are:

• The normalisation of variable, in Normalising the variable

• The ranking of variable, in Computing the ranking

• The elementary statistic computation, in Computing the elementary statistic

• The quantile estimation, in The quantile computation

• The correlation matrix determination, in Correlation matrix

• 2.4.1. Normalising the variable
• 2.4.2. Computing the ranking
• 2.4.3. Computing the elementary statistic
  • 2.4.3.1. Specific case of vectors
• 2.4.4. The quantile computation
  • 2.4.4.1. computeQuantile
  • 2.4.4.2. \(\alpha\)-quantile
  • 2.4.4.3. Wilks-quantile computation
• 2.4.5. Correlation matrix
  • 2.4.5.1. Special case of vector