For plotting mirrored similarity matrices, it would be nice to be able to plot only half, i.e. the diagonal. Because of wanting to cluster as well, if you would set some values to NA in the input_matrix the internal hclust() call on dist(input_matrix) would error because of NA values.
I love pheatmap, it would be silly to need another package just for this.
Hi,
For plotting mirrored similarity matrices, it would be nice to be able to plot only half, i.e. the diagonal. Because of wanting to cluster as well, if you would set some values to NA in the input_matrix the internal hclust() call on dist(input_matrix) would error because of NA values.
I love pheatmap, it would be silly to need another package just for this.