On 2/06/2009 10:17 AM, Vincent Arel wrote:
> Your python-like example is also quite helpful.
It is not "python-like". Apart from the "..." in the initial data
"vectors", it is executable Python code.
> As I understand it, you
> basically implement Igor's suggestion of running loops on the vectors.
More or less :-) BTW, I miscoded; see below.
> I
> should be able to do that quite easily. Most importantly though, it appears
> that I need to revise my understanding of the division of labour between
> sqlite and R.
>
> Thanks a lot for your help, and have a great week!
You're welcome ... you too.
[big snip]
>>
>> var1 = ["ALB", "CAN", "DZA", ...]
>> var2 = ["ALB", "CAN", "DZA", ...]
>> var3 = ["1961", "1962", "1963",...]
>> # get a list of unique country codes, in sorted order
>> countries = list(set(var1 + var2))
>> countries.sort()
>> # convert years to integer, find range
>> var3int = [int(y) for y in var3]
>> firsty = min(var3int)
>> lasty = max(var3int)
>> year_range = range(firsty, lasty + 1)
>> # do the business
>> id = 0
following works very hard to throw away half of your results :-(
>> ncountries = len(countries)
>> for i in range(ncountries - 1):
>> for j in range(i + 1, ncountries):
>> assert countries[i] != countries[j]
>> id += 1
>> for year in year_range:
>> print id, countries[i], countries[j], year
should be:
for c1 in countries:
for c2 in countries:
if c2 == c1: continue
id += 1
for year in year_range:
print id, c1, c2, year
and sorting the countries is not essential, but maybe helpful when
visually checking the output.
Cheers,
John
_______________________________________________
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
