(Don't top-post. Put your new remarks after whatever you're quoting)
On 09/26/2011 08:26 PM, questions anon wrote:
Thanks for your feedback. I might be asking the wrong questions.
I have written a number of scripts that do various things for each step:
1. get data (e.g. Dec-jan-feb, only januarys, all data)
2. summary stats(e.g. max, min, frequency)
3. plot data (e.g. plot with time, plots subplots, plot with particular
projection)
I would like to then be able to choose which script for each step and
somehow feed the data/variables written in earlier functions through to then
be able to run something like:
def main():
getdataDJF()
calculatemean()
plotmean()
main()
Is this possible? Is this the correct way to do things?
My main problem is not being able to feed the output from one function into
the next.
Thanks
Sarah
If multiple functions need to share data, it should be done by passing
parameters, and returning results. If you can understand how to do that
within a single source file, you'll be ready to extend the concept to
multiple source files.
But if you just want to skip to a solution, without worrying about
whether it makes sense, read on.
You have multiple files, how do they relate to each other? What is the
kind of data you're trying to share?
I'll make a crazy guess, since you still give no clue.
Perhaps each of your multiple files, called external1.py, external2.py
defines three functions, called getdataDJF(), calculatemean(), and
plotmean(). They're defined as follows:
getdataDJF - takes arg1, arg2 and arg3 as parameters, and returns
results as (res1, res2)
calculatemean - takes arg1 and arg2 as parameters, and returns mydata
as result
plotmean - takes mydata and device as parameters, and doesn't return
anything
So your script could do a conditional import to get various
implementations of the functions.
import sys
if (something):
from external1 import getdataDJF, calculateman, plotmean
else:
from external2 import getdataDJF, calculateman, plotmean
def main(argv):
(res1, res2) = getdataDJF(argv[1], argv[2], argv[3])
buffer = calculatemean(res1, res2)
plotmean(buffer, argv[4])
main(sys.argv, "printer")
Another approach if this guess is wrong: use classes. Keep all the
data as attributes of a class instance, and the methods can share data
by stuffing it in those attributes.
--
DaveA
_______________________________________________
Tutor maillist - [email protected]
To unsubscribe or change subscription options:
http://mail.python.org/mailman/listinfo/tutor
