Skip to content

Data Frame Column Name Attribute

3 messages · thomas mann, David Winsemius, William Dunlap

thomas mann
# 
I am attempting to add a calculated column to a data frame.  Basically,
adding a column called "newcol2" which are the stock closing prices from 1
day to the next.

The one little hang up is the name of the column.  There seems to be an
additional data column name included in the attributes (dimnames?).  So
when i run HEAD(DATAFRAMENAME) i get the column name = "Open".   but you
can see below... it is called "newcol2 with some attribute called open.

I did a lot of google searches but really could not find what i was looking
for.  Any help is greatly appreciated.  Thank you!

tdata[["newcol2"]] <- rbind(0,apply(tdata["Open"],2,diff))
'data.frame': 7505 obs. of  8 variables:
 $ Date   : int  19860709 19860710 19860711 19860714 19860715 19860716
19860717 19860718 19860721 19860722 ...
 $ Open   : num  9.14 9.45 9.3 9.08 8.71 ...
 $ High   : num  9.45 9.51 9.35 9.08 8.71 ...
 $ Low    : num  9.03 9.14 9.03 8.65 7.96 ...
 $ Close  : num  9.4 9.35 9.03 8.76 8.5 ...
 $ Volume : int  332719 180049 234212 218772 605427 339507 306866 418593
94880 119332 ...
 $ OpenInt: num [1:7505, 1] 0 0.317 -0.155 -0.223 -0.367 ...
  ..- attr(*, "dimnames")=List of 2
  .. ..$ : NULL
  .. ..$ : chr "Open"
* $ newcol2:* num [1:7505, 1] 0 0.317 -0.155 -0.223 -0.367 ...
  ..- attr(*, "dimnames")=List of 2
  .. ..$ : NULL
*  .. ..$ : chr "Open"*
David Winsemius
# 
The attribute comes from the name of the argument to `apply`
The `rbind` function is returning a matrix and `[[<-. data.frame` is adding it to the tdata object. Matrices have dimname attributes. It appears to me that you have also earlier attempted (successfully) to add a column named "OpenInt" higher up in your code that you have not shown us.

If you want to get rid of this column use `[<-.data.frame` with a NULL argument

tdata['OpenInt'] <- NULL


If you start again with neither of these columns, I would suggest this instead:

tdata[["newcol2"]] <- c( 0, apply(tdata["Open"],2,diff ) )

The valued returned should be an undimensioned object and will not get bound to the dataframe as a matrix.
William Dunlap
# 
You could use transform() instead of [[<- to add columns to your data.frame
so the new columns get transformed they way they do when given to the
data.frame function itself.  E.g.,
'data.frame':   5 obs. of  3 variables:
 $ X   : int  1 2 3 4 5
 $ Y   : int  11 12 13 14 15
 $ NewZ: int  12 14 16 18 20
c("NewZ1","NewZ2")))))
'data.frame':   5 obs. of  4 variables:
 $ X      : int  1 2 3 4 5
 $ Y      : int  11 12 13 14 15
 $ Z.NewZ1: int  1 2 3 4 5
 $ Z.NewZ2: int  11 12 13 14 15



Bill Dunlap
TIBCO Software
wdunlap tibco.com

On Sat, Apr 23, 2016 at 8:59 AM, thomas mann <twiniverse2000 at gmail.com>
wrote: