Should I reformat my csv data that I am ingesting?

[bpiela]

I have a csv file that is created by a shell script on a Linux server and runs every minute.  I am running a forwarder on the server to send the data to splunk.  The csv file has a header line containing the field names.  Some of the data has units along with the number.  What I mean is if the number is a percent, the number might look like "10%".  If a number is measured in microseconds, the number might look like "10us".  Is that the best way to ingest the data?  Should I remove the units?  Does Splunk see "10us" as a number?   Thanks!

[richgalloway]

Splunk will interpret "10us" as a string because it contains characters that don't belong in a number (sign, digits, decimal separator).  To ensure Splunk treats numbers as numbers, remove any units before ingestion, if possible.