prep_batch_ranges() causes the remainer to be dropped

[sgsutcliffe]

Problem

In the matrix_splitter class the function for dividing up a matrix into batches prep_batch_ranges does not update the ranges if their is a remainer, thus droping the samples during performing batches.

Proposed Solution

Add the num_lines to last batch range so that all samples are included.
e.g.

def prep_batch_ranges(num_lines, batch_size):
    num_batches = int(num_lines / batch_size)
    rem = batch_size % num_batches
    ranges = []
    for i in range(0,num_batches):
        ranges.append((i*batch_size,i*batch_size+batch_size))
    if rem != 0:
        r = ranges[-1]
        r = list(r)
        r[1] = num_lines
        ranges[-1] = r
    return ranges

print(prep_batch_ranges(32,5))

Note

This will impact the tests that are being created in the PR #7