simple first emacs script

[Tom]

I am about 2/3 way through the introduction to emacs lisp, and thought 
it would help my learning if I wrote tried writing a simple script to do 
something useful to me with the lessons I have covered so far so far.  I 
should point out I am surgeon not a computer scientist and know nothing 
more about programming than what is contained in the introduction to 
emacs lisp.
I end up with this:

(defun nirs-data-clean (number-of-channels)
"Cleans the output files from the INVOS monitor removing all columns 
apart from StO2 values, and the date and time stamps. Sto2 columns to be 
retained are specified by the NUMBER-OF-CHANNELS variable.  It assumes 
that channels are always used in numerical order, i.e. channel 1 always, 
then 2, then 3 then 4."
(interactive "nEnter number of channels (1-4): ")
(barf-if-buffer-read-only)
(require 'csv-mode)
(push-mark (point-max))
(goto-char (point-min))
(if (or (eq number-of-channels 1)
        (eq number-of-channels 2)
        (eq number-of-channels 3)
        (eq number-of-channels 4))
    (or (when (eq number-of-channels 1)
	  (csv-kill-fields '(4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 
23 24 25 26 27 28 29 30 31 32 33 34) (point) (mark)))
        (when (eq number-of-channels 2)
	  (csv-kill-fields '(4 5 6 7 8 9 11 12 13 14 15 16 17 18 19 20 21 22 23 
24 25 26 27 28 29 30 31 32 33 34) (point) (mark)))
        (when (eq number-of-channels 3)
	  (csv-kill-fields '(4 5 6 7 8 9 11 12 13 14 15 16 18 19 20 21 22 23 24 
25 26 27 28 29 30 31 32 33 34) (point) (mark)))
        (when (eq number-of-channels 4)
	  (csv-kill-fields '(4 5 6 7 8 9 11 12 13 14 15 16 18 19 20 21 22 23 25 
26 27 28 29 30 31 32 33 34) (point) (mark))
	  (message "%s is not a valid number of channels.  Please enter a 
number between 1 and 4." number-of-channels))))
(when (y-or-n-p "Replace 0 StO$_2$ values with na: ")
    (progn (goto-char (point-min))
           (while (re-search-forward "[^[:digit:]][0][^[:digit:]]" nil t)
             (replace-match "na ")))))

The idea is to automatically clean unwanted columns from csv (space 
separated) data file and ask me if I want to replace 0 values with na. 
I appreciate this is very messy, and I should be able to specify ranges 
of field rather than listing every single one (but I couldn't get that 
to work), but it works kind of.  When I evaluate it and run it for the 
first time it works, but if I undo the changes to the file and run it 
again it often only takes out the 4th column, if I reopen the file or 
re-evaluate the script it usually works again but I can't seem to work 
the pattern of working and not working - as I can't see the pattern I 
can't learn where I am going wrong.

I have attached a small sample data file with 4 channels being used to 
illustrate the kind of file this is supposed to work on.

I assume I have made an obvious mistake but I don't have the experience 
to know what it is.  Any general comments about simple/better ways to 
clean this script would be welcome, for example would I be better making 
the regexp replace a yes/no arg to the function so if I call this script 
from another script I can specify a y/n argument (I don't know how to do 
this anyway though).

Tom

101108N.R14
Description: Text document