Fisher test using R for data in Splunk

melonman — Thu, 23 Apr 2015 12:07:31 GMT

I want to perform fisher test on data in Splunk. The table passed to R looks like this:

index=summary stats=conversion_rate 
| table _time purchase view
| fields - _time

purchase view
-------- ----
     747 2463
     537 2545

I want to to the fisher.t(input), to compare 1st line and 2nd line, and get p-value as a result.

but I got stuck at a very basic step, creating matrix from data passed from Splunk.
Could anyone give me a sample for this?

Re: Fisher test using R for data in Splunk

clongo01 — Thu, 23 Apr 2015 18:45:14 GMT

Hello. You will have to format your data in a table at then end of your search. The R app will automatically assign that table as the input to your R command. It would look something like this.

index=summary stats=conversion_rate | table purchase view | r "
result = fisher.test(input)
output = data.frame(Result=c(result))
"

Your output will be two columns because it is capturing the confidence bounds as well and the p-value. If you want to return just the p-value from the fisher test change the last line to be output = data.frame(Result=c(result$p.value)).

Hope this helps!

topic Re: Fisher test using R for data in Splunk in Splunk Search

Fisher test using R for data in Splunk

Re: Fisher test using R for data in Splunk