How do I get the minimum or maximum of two values for each row in a data.table?

Asked Apr 20 '17 at 15:15

Active Apr 20 '17 at 15:15

Viewed 28 times

I have a list of standard errors and standard deviations by date, like so:

date        avg_stdev   avg_sterr
2016-11-25  0.02345404  0.009561593
2016-11-27  0.02098070  0.009600404
2016-11-28  0.02098070  0.009600404
2016-11-29  0.02098070  0.009600404
2016-11-30  0.02009983  0.008637851
2016-12-03  0.01793612  0.008303243

I want to create two new columns that are the higher and the lower of those. (So in each of these particular rows, avg_stlow would equal avg_sterr and avg_sthigh would be avg_stdev, but the values will cross later on.)

I tried doing dt[,avg_stlow := min(avg_stdev, avg_sterr)], but this just gives me the minimum value from the columns. I also tried dt[,avg_stlow := if(avg_stdev > avg_sterr) avg_sterr else avg_stdev], but this seems to determine which column is less (by some means) and use that column throughout. How can I get it to evaluate based on the value for each row?

asked Apr 20 '17 at 15:15

futuraprime

5,468
7
38
58

2

It's the strangely-named `pmin` and `pmax` you need, I guess. – Frank Apr 20 '17 at 15:17
Yes! This is it. (Stick it in as an answer and I'll mark it.) – futuraprime Apr 20 '17 at 15:19
Glad it works. I think this linked Q&A covers it. – Frank Apr 20 '17 at 15:24

How do I get the minimum or maximum of two values for each row in a data.table?

0 Answers0