CODEWITHSUNDEEP
rxts
E.D.
E.D.

Reputation: 329

Add factor column to xts object

Using the unique command, I can easily get a list of unique tickers from my raw daily stock data. Which results in:

my.table <- unique(my.frame1[,5])
> my.table
 [1] NPN BIL CFR IMP FSR SHF SHP REI INP OML REM ABL AGL SAB WHL BTI MMI RMI PIK TRU INL SPP CLS
[24] SBK PPC IPL SOL ASA RMH MTN ANG EXX DSY NED SLM AVI KIO VOD GFI APN
40 Levels: ABL AGL ANG APN ASA AVI BIL BTI CFR CLS DSY EXX FSR GFI IMP INL INP IPL KIO ... WHL

However, I now loop through the list to do a few basic calcs. I am trying to assign the ticker code to each trading day, but the observation in the list do not interact with my data as expected:

> code <- my.table[2]
> code
[1] BIL
40 Levels: ABL AGL ANG APN ASA AVI BIL BTI CFR CLS DSY EXX FSR GFI IMP INL INP IPL KIO ... WHL 

> my.daily <- my.daily[,3:6]
> my.daily[,1] <- code
> my.daily
           my.xts...1..Low my.xts...1..Close     VWAP SIDE
2011-08-31               7             23765 23744.30    1
2011-09-08               7             22876 23056.72    1
2012-02-14               7             25050 25230.52   -1
2012-07-25               7             23480 23591.01   -1
2013-05-12               7             26818 26737.65   -1

Now I would have expected the first column in my xts object to be "BIL", as per the value of code above.

What am I missing please?

Upvotes: 0

Views: 750

Answers (1)

agstudy
agstudy

Reputation: 121626

Your xts is a numeric matrix. You can't have multi-types (numeric and factor) in a matrix. The factor is coerced to a factor (as already commented) to have a single type(numeric here).

As a workaround, you can firstly coerce your xts object to a data.frame.

my.daily <- as.data.frame(my.daily)
my.daily[,1] <- code

But why are you trying to do this? You loose all the power and efficiency of xts to deal with time series.

Upvotes: 2

Related Questions