How to use acast correctly on R data frame?

Asked Oct 05 '17 at 21:02

Active Oct 05 '17 at 21:07

Viewed 124 times

I have an R data frame in this form with thousands of records:

account_id             timestamp    days_since_launch   region  used
    1263562             6/23/16           5               SFO    57
    3821750             6/24/16           11              ATL    10

I want to convert it into a form, where the column names are:

account_id    timestamp  days_since_launch  SFO ATL SEA NYC SLC

Note: In here the columns SFO, ATL, SEA, NYC, SLC are the unique entries of region in the original dataset. I want the entries in these region columns to have the corresponding numeric entry in used.

How can this be done in R?

I tried as.data.frame(acast(inputDF, account_id + timestamp ~ region, value.var="used"))

The issue I have with this is: it forms a column name called account_id_timestamp with the account id's and timestamps concatenated like 1263562_6/23/16 . It also gives a warning: "Aggregation function missing: defaulting to length"

edited Oct 05 '17 at 21:07

Psidom

209,562
33
339
356

asked Oct 05 '17 at 21:02

hearse

Please provide a reproducible example, what you've tried and where you feel you fall short. – Roman Luštrik Oct 05 '17 at 21:02
`acast` is from reshape2, not dplyr. – Frank Oct 05 '17 at 21:04
I have included what I tried. I am not restricting the answer to be only through acast. Any other function could be used as part of the accepted answer-as long as it works. – hearse Oct 05 '17 at 21:08
2

Does `spread(df, region, used)` work for you? – Psidom Oct 05 '17 at 21:13
It works like a charm – hearse Oct 05 '17 at 21:14

How to use acast correctly on R data frame?

0 Answers0