Suppose I have a data set with a metric, let's say for example, it contains the average # of stamps licked per day by a person in a bulk mailer operation with an unknown, inconsistent # of stamp lickers.
ideally the data would be like this:
StampLicker Date Produced
Rita , 8/1/2018 , 15
Henry , 8/1/2018 , 14
Stefan , 8/1/2018 , 15
Rita , 8/2/2018 ,16
Henry , 8/2/2018 ,14
Stefan , 8/2/82018 ,15
Amy , 8/3/2018 , 16
Henry , 8/3/2018 , 14
Stefan , 8/3/82018 , 15
But unfortunately I have it like this:
Date AMY RITA HENRY STEFAN
8/1/2018 , ,15 ,14,15
8/2/2018, ,16,14,15
8/3/2018,16 , 14,15
These have run for a while, perhaps years. Stamp lickers have come and gone. I'm interested in seeing metrics for most recent day as well as historical average to produce a chart like this:
Chart for most recent day
AMY RITA HENRY STEFAN
Amy X NA -2 1
Rita NA X 2 1
Henry -2 -2 X -1
Stefan -1 -1 1 X
and perhaps a trend line indicating success or failure of the stamp licking training program when a new stamp licker is hired. Any thoughts on how to compute that table without doing|eval HenryVSAmy=Henry-Amy and so on because I don't know how many/what the columns will be in advance necessarily. It's similar to the XYSeries function but not.
... View more