Method ProjectColToTable

tensor/stats/pca/pca.go:195–230

ProjectColToTable projects values from the given column of given table (via IndexView) onto the given set of eigenvectors (idxs, 0 = largest eigenvalue, 1 = next, etc), and stores results along with labels from column labNm into results table. Must have already called PCA() method.

(projections *table.Table, ix *table.IndexView, column, labNm string, idxs []int)

// and stores results along with labels from column labNm into results table.
// Must have already called PCA() method.
func (pa *PCA) ProjectColToTable(projections *table.Table, ix *table.IndexView, column, labNm string, idxs []int) error {
	_, err := ix.Table.ColumnByNameTry(column)
	if err != nil {
		return err
	}
	if pa.Vectors == nil {
		return fmt.Errorf("PCA.ProjectCol Vectors are nil -- must call PCA first")
	}
	rows := ix.Len()
	projections.DeleteAll()
	pcolSt := 0
	if labNm != "" {
		projections.AddStringColumn(labNm)
		pcolSt = 1
	}
	for _, idx := range idxs {
		projections.AddFloat64Column(fmt.Sprintf("Projection%v", idx))
	}
	projections.SetNumRows(rows)

	for ii, idx := range idxs {
		pcol := projections.Columns[pcolSt+ii].(*tensor.Float64)
		pa.ProjectCol(&pcol.Values, ix, column, idx)
	}

	if labNm != "" {
		lcol, err := ix.Table.ColumnByNameTry(labNm)
		if err == nil {
			plcol := projections.Columns[0]
			for row := 0; row < rows; row++ {
				plcol.SetString1D(row, lcol.String1D(row))
			}
		}
	}
	return nil
}