Applying SetNames to a list of data frames

Question

I'm running into an issue applying new names to a list of data frames. I'm using quantmod to pull stock data, and then calculating the 7-Day moving average in this example. I can create the new columns within the list of data frames, but when I go to rename them using lapply and setNames it is only returning the newly renamed column and not any of the old data in each data frame.

require(quantmod)
require(zoo)

# Select Symbols
symbols <- c('AAPL','GOOG')

# Set start Date
start_date <- '2017-01-01'

# Get data and put data xts' into a list. Create empty list and then loop through to add all symbol data
stocks <- list()

for (i in 1:length(symbols)) {
  stocks[[i]] <- getSymbols(symbols[i], src = 'google', from = start_date, auto.assign = FALSE)
}

##### Create the 7 day moving average for each stock in the stocks list #####

stocks <- lapply(stocks, function(x) cbind(x, rollmean(x[,4], 7, align = "right")))

Sample Output:

[[1]]
               AAPL.Open AAPL.High AAPL.Low AAPL.Close AAPL.Volume AAPL.Close.1
    2017-01-03    115.80    116.33   114.76     116.15    28781865           NA
    2017-01-04    115.85    116.51   115.75     116.02    21118116           NA
    2017-01-05    115.92    116.86   115.81     116.61    22193587           NA
    2017-01-06    116.78    118.16   116.47     117.91    31751900           NA
    2017-01-09    117.95    119.43   117.94     118.99    33561948           NA
    2017-01-10    118.77    119.38   118.30     119.11    24462051           NA
    2017-01-11    118.74    119.93   118.60     119.75    27588593     117.7914
    2017-01-12    118.90    119.30   118.21     119.25    27086220     118.2343
    2017-01-13    119.11    119.62   118.81     119.04    26111948     118.6657

    [[2]]
               GOOG.Open GOOG.High GOOG.Low GOOG.Close GOOG.Volume GOOG.Close.1
    2017-01-03    778.81    789.63   775.80     786.14     1657268           NA
    2017-01-04    788.36    791.34   783.16     786.90     1072958           NA
    2017-01-05    786.08    794.48   785.02     794.02     1335167           NA
    2017-01-06    795.26    807.90   792.20     806.15     1640170           NA
    2017-01-09    806.40    809.97   802.83     806.65     1274645           NA
    2017-01-10    807.86    809.13   803.51     804.79     1176780           NA
    2017-01-11    805.00    808.15   801.37     807.91     1065936     798.9371
    2017-01-12    807.14    807.39   799.17     806.36     1353057     801.8257
    2017-01-13    807.48    811.22   806.69     807.88     1099215     804.8229

I would like to change the "AAPL.Close.1" and "GOOG.Close.1" to say "AAPL.Close.7.Day.MA" and "GOOG.Close.7.Day.MA" respectively (for however many symbols that I choose at the top).

The closest that I've gotten is:

stocks <- lapply(stocks[], function(x) setNames(x[,6], paste0(names(x[,4]), ".7.Day.MA")))

This is correctly naming the new columns, but now my stocks list only contains that single column for each ticker:

[[1]]
           AAPL.Close.7.Day.MA
2017-01-03                  NA
2017-01-04                  NA
2017-01-05                  NA
2017-01-06                  NA
2017-01-09                  NA
2017-01-10                  NA
2017-01-11            117.7914
2017-01-12            118.2343
2017-01-13            118.6657

[[2]]
           GOOG.Close.7.Day.MA
2017-01-03                  NA
2017-01-04                  NA
2017-01-05                  NA
2017-01-06                  NA
2017-01-09                  NA
2017-01-10                  NA
2017-01-11            798.9371
2017-01-12            801.8257
2017-01-13            804.8229

Why is the setNames function removing the original columns?

amonk · Accepted Answer · 2017-06-26T10:42:13.597

Almost there:

N = 10 #number of pseudorandom numbers
df1 <- data.frame(a=runif(N),b=sample(N))#1st data frame
df2 <- data.frame(c=rnorm(N),google=df1$b^2,e=df1$a^3)#2nd data frame
stocks<-list(df1,df2)# create the list

lapply(stocks,names) # get the names of each list element (data.frame) 
[[1]]
[1] "a" "b"

[[2]]
[1] "c"      "google" "e"

Since we are using a function we need to use the <<- in order to overwrite the initial object stocks.

lapply(seq_along(1:length(stocks)),function(x) names(stocks[[x]])<<-gsub(pattern = "google",replacement = "google2",x=names(stocks[[x]])))#replacing the string google 

[[1]]
[1] "a" "b"

[[2]]
[1] "c"       "google2" "e"

Additionally (verification) stocks contains the new names:

> stocks
[[1]]
            a  b
 1  0.73826897  3
 2  0.35627664  8
 3  0.89060134  7
 4  0.72629312 10
 5  0.97069742  4  
 6  0.12530931  2
 7  0.65744257  9
 8  0.06218019  1
 9  0.67322891  6
 10 0.66128204  5

 [[2]]
             c google2           e
 1  -0.5272267       9 0.402386917
 2   0.6993945      64 0.045223278
 3   0.3707304      49 0.706398932
 4  -0.2371541     100 0.383120861
 5   1.5073834      16 0.914643019
 6   0.4098821       4 0.001967660
 7  -0.3014211      81 0.284166886
 8   0.3248919       1 0.000240412
 9   1.2757740      36 0.305132358
 10  1.5938208      25 0.289174620

Thanks! gsub looks to work well here. I'm still not sure why my original code removed all of the columns that weren't affected though. Can you explain why? — Jason J, Jun 27 '17 at 01:21
Check this out: https://stackoverflow.com/questions/6081439/changing-column-names-of-a-data-frame-in-r — amonk, Jun 27 '17 at 06:51

Applying SetNames to a list of data frames

1 Answers1