Encode with Correlation-Ordered Grid

Usage

encode_corr_grid(x, order, rows = NULL, cols = NULL)

Arguments

x: Numeric feature vector (length p)
order: Integer vector from fit_corr_order
rows: Number of rows in the output grid. Defaults to ceiling(sqrt(length(x))).
cols: Number of columns in the output grid. Defaults to ceiling(length(x) / rows).

Value

A rows × cols numeric matrix

References

Sharma A et al. (2019). DeepInsight: A methodology to transform a non-image data to an image for convolution neural network architecture. Scientific Reports, 9, 11399.

Examples

encode_corr_grid(1:6, order = c(6, 5, 4, 3, 2, 1), rows = 2, cols = 3)
#>      [,1] [,2] [,3]
#> [1,]    6    5    4
#> [2,]    3    2    1