exit a python data function early

[David Katz]

At the start of my python data function, I validate the input. If the input is not valid, what is the best way to return to spotfire immediately?

[Gaia Paolini]

What I tend to do is:

• create an 'empty_return' function to call if the input is invalid or there is an early termination
• this function returns exactly the same output parameters as needed, but empty versions of them
• I add a message output string to the output parameters, so it can be saved in a document property and displayed.

Other very useful tips on Python data functions in Spotfire in this Community article:

https://community.spotfire.com/s/article/Tips-and-Tricks-for-Working-with-Python-in-Spotfire

Simple example: I put the main code in a function so it could be exited early in different places if needed.

In this case there is an integer input parameter my_parm and three output parameters: two data frames and a message string.

import pandas as pdimport numpy as np def empty_return(msg): df1=pd.DataFrame({'errcol':[np.nan]}) df2=pd.DataFrame({'errcol':[np.nan]}) return (df1,df2,msg) def main(parm): if parm==0: return empty_return('parameter was zero') #otherwise normal execution aa=[1,2,3] bb=['A','B','C'] cc=['a','b','c'] df1=pd.DataFrame({'col1':aa,'col2':bb}) df2=pd.DataFrame({'col1':aa,'col2':cc}) message='all good' return (df1,df2,message) df1,df2,message=main(my_parm) 

[David Katz]

Thanks Gaia.

 

This is a good solution. I was hesitant to use it because my data function has a long list of parameters, and I was trying to avoid this:

 

main(x,y,z,a,b,c...)

 

But your endorsement of 'main' led me to consider this - use the tip in the link you provided to load the Spotfire parameters directly into main():

 

def main(debug):

if debug = True:

 

main(False) within SF

main(True) to debug in emacs or...

 

Does this make sense to you?

 

*David Katz*, TIBCO Data Science

 

 

1.541.324.7417