rkanters.txt

2004.2.20
Hi Dan,

I had the need to convert maidenhead to long/lat coordinates for a
faculty member here who maintains a database with disasters around the
world, so I searhced the web and found your python scripts. I am not a
python programmer, but was able to convert your scripts into a set of
Visual Basic functions for use in Excel, and into Coldfusion MX for a
web based service.

Since divmod is a particularly peculiar python function not available
in VB type languages (I think), I rewrote the lonlat2maiden function.
The file is enclosed (it is a little bit shorter, but I am not sure
about the execution time differences). Notice that I also changed the
algorithm a bit in that I create all the characters within the same
loop, by first dividing out the first level grid sizes and setting the
loop up to generate the character pair first.

About the maiden2lonlat script. I think you would get a bit better
performance (specially if you use the procedure a lot) if you were to
expand the f function in the last loop. (I enclosed the mh2ll file
also). That f function is indeed a stroke of genius. I had to wreck my
mind for quite a while to figure it out.

I noticed some conversion round-off errors when going from
lonlat2maiden. I think the culprit is that for the last division the
result is truncated (like all the other ones) where it probably should
have been rounded. For instance I get:

echo PL04KE | maiden2lonlat | lonlat2maiden
PL04JE

and the remainder can be seen:
echo PL04KE | maiden2lonlat | lonlat2maiden 22
PL04JE90XA90XA90XA90KC

Thanks for sharing your code!!
Cheers,
René

PS I hope the email stayed small enough :-)

Hi Dan,

Thanks for the reply.

I know how you feel when you put a piece of code out there and years
later get messages about it :-).

I made what I consider my final modifications to two versions of the
mh2ll (so I won't bother you again). The mh2ll3 is the implementation I
use myself, while mh2ll2 is one where I made the calculation of the
resolution of the grid by doing the continued division (i.e.,
algorithmically the inverse of what is done in the ll2mh procedure). I
opted for that since it is computationally less expensive (no
background logarithms needed for the power raising functions). I added
some extra comments to the mh2ll3 version.

I also like short pieces of code (Occam's razor), and since I believe
in garbage in--garbage out, I don't do much error checking (which will
come back to haunt me some time).

I would like to take you up on the offer of putting my email onto your
page as "one user's improvements to there programs", only if you agree
that they are improvements...

Cheers,
René

René P.F Kanters, Ph.D.
Director of Computer Assisted Science Education
Department of Chemistry
University of Richmond, Virginia 23173
Phone: (804) 287-6873
Fax: (804) 287-1897
Email: rkanters@richmond.edu
Web: http://www.richmond.edu/~rkanters