The ranklet transform, gives output which ranges from [-1,1] but while calculating GLCM features values must be positive. So how to calculate GLCM feature from vertical, horizontal and diagonal ranklet images.
There are numerous options. The first, of course, is to simply normalize the data using the formula (ranklet + 1)/2. However, you can also analyze the frequency of each ranklet value and use that to either evenly spread your ranklets from 0 to 1 or create logical buckets.