topic Re: Check for pure numeric field values or alphanumeric in Splunk Search

Check for pure numeric field values or alphanumeric

AshimaE — Thu, 06 Jul 2017 09:51:29 GMT

I have extracted a field using regex having the name push. I now have to classify all my rows into two categories based on whether push is all numeric or contains some alphabets also. Let it be divided into classes pure and mixed and given in a field class.
Any suggestions on how to do this since the numbers involved are too huge going upto 15-20 characters.How to fill in the class field for this statement.

Re: Check for pure numeric field values or alphanumeric

gcusello — Thu, 06 Jul 2017 09:59:43 GMT

Hi AshimaE,
try something like this

your_search
| rex field=push "(?<type_num>[0-9]*)"
| eval type=if(type_num=*,"Numeric","Alfanumeric")
| stats count by type

Bye.
Giuseppe

Re: Check for pure numeric field values or alphanumeric

niketn — Thu, 06 Jul 2017 10:51:20 GMT

Hi Ashima, you can use isnum() evaluation function. Following is run-anywhere search.

| makeresults
| eval push="TestData123"
| eval class=if(isnum(push),"numeric","alpha")

PS: Above one will treat negative and decimal numbers also as numeric.

Re: Check for pure numeric field values or alphanumeric

niketn — Fri, 07 Jul 2017 17:05:08 GMT

@AshimaE, can you please test and confirm whether the suggested change works for you?