I have a set of test results for a list of products over a period of time. I need to create a Alteryx workflow that can recognize the signal of the numeric data - up, down, flat - this will lead to a pattern recognition on the data to evaluate trends. The set sample is as follow and I am looking to have the "Signal" column filled:
Thank you.
As the timeframes between records are not consistent, you would either need to aggregate or generate to create a time series. If you are simply trying to flag whether each record is up/down/flat from the last record, then take a look at the Multi-Row Formula tool.
IF IsNull([Row-1:TestResult])
THEN 'flat'
ELSEIF [Row-1:TestResult]<[TestResult]
THEN 'up'
ELSEIF [Row-1:TestResult]>[TestResult]
THEN 'down'
ELSE 'flat'
ENDIF
Kane
Edit: changed second up to down...