Aligning data

[DeaD_EyE]

Numpy is usually for Numbers and not for strings used.
You can do this with Python itself:

from itertools import chain
from operator import itemgetter

# your data
sample_data = [
        ('sample001a.com', -12, 'sample002b.com', -16, 'sample002c.com', -18),
        ('sample002a.com', -14, 'sample003b.com', -16, 'sample004c.com', -22),
        ('sample003a.com', -16, 'sample004b.com', -20, 'sample005c.com', -16),
        ('sample004a.com', -18, 'sample005b.com', -14),
        ('sample005a.com', -12),
    ]

def sort_column(row):
    # ('sample003a.com', -16, 'sample004b.com', -20, 'sample005c.com', -16)
    # this have to be sorted
    # grouping host and value together
    group_by_two = zip(*[iter(row)]*2)
    # you can do this instead with indexing
    sorted_groups = sorted(group_by_two, key=itemgetter(1))
    # group_by_two is the iterator
    # itemgetter get the second element from iterable (1)
    # the list is sorted by this value, the second value
    # now need to flatten the iterable
    # retunring it as imuatble tuple
    return tuple(chain.from_iterable(sorted_groups))


for row in sample_data:
    print(sort_column(row))

How you get the data, is you task. I put an example data set in the top as a nested list.