CODEWITHSUNDEEP
rreshape
user2238328
user2238328

Reputation: 259

R reshape package

I am attempting to use R's "reshape" package to alter a data frame I have, pasted below. Basically, I would like to create a new data frame which concatenates variables V2 and V3 into columns for all possible combinations of V2 and V3, with rows as date. I have attempted using reshape's cast function, however I am unable to get the Date to appear in the rows. The variables are concatenated in column names as I want, but the Date is not included.

This is the data frame std:

V2      V3   V5       Date
AUS     CR   15344    2000-01-01
ALI     NG   3952     2000-01-01
EUR     CR   19296    2000-01-01
AUS     MO   5826     2000-01-01

When I apply cast(std,Date~V2~V3,value="V5"), I obtain a frame as shown below, however I am unable to get the Date column to show as well. I have done lots of trial and error with cast and melt to no avail. Any suggestions would be appreciated

AUS.CR  ALI.NG  EUR.CR
15344   3952    19296
20108   4000    18568

Upvotes: 0

Views: 252

Answers (1)

mathematical.coffee
mathematical.coffee

Reputation: 56905

You could use the package reshape (function cast) or reshape2 with dcast:

x <- read.table(header=T,text='V2      V3   V5       Date
AUS     CR   15344    2000-01-01
ALI     NG   3952     2000-01-01
EUR     CR   19296    2000-01-01
AUS     MO   5826     2000-01-01')

With reshape and cast:

library(reshape)
cast(x, Date ~ V2 + V3, value='V5')
#         Date ALI_NG AUS_CR AUS_MO EUR_CR
# 1 2000-01-01   3952  15344   5826  19296

With reshape2 and dcast (same result)

library(reshape2)
dcast(x, Date ~ V2 + V3, value.var='V5')

Upvotes: 3

Related Questions