Same code in different environments render different results in R

Question

The following code renders a certain response to a query:

library(quantmod)

# Load ticker data from 2020-01-01 till 2021-02-02
t <- c("NKLA", "MPNGF", "RMO", "JD", "MSFT")
getSymbols.yahoo(t, auto.assign = TRUE, env = globalenv(), 
                 from = "2020-01-01", to = "2021-02-02")

# Close all Internet connections as a precaution
closeAllConnections()

# Find xts objects
x <- names(which(unlist(eapply(.GlobalEnv, is.xts))))

# Convert xts to data.frame
for (i in seq_along(x)) {
  assign(x[i], fortify.zoo(get(x[i])))
}

# The query
sapply(mget(x), names)

# The rendering
     NKLA            MPNGF            MSFT            JD            RMO           
[1,] "Index"         "Index"          "Index"         "Index"       "Index"       
[2,] "NKLA.Open"     "MPNGF.Open"     "MSFT.Open"     "JD.Open"     "RMO.Open"    
[3,] "NKLA.High"     "MPNGF.High"     "MSFT.High"     "JD.High"     "RMO.High"    
[4,] "NKLA.Low"      "MPNGF.Low"      "MSFT.Low"      "JD.Low"      "RMO.Low"     
[5,] "NKLA.Close"    "MPNGF.Close"    "MSFT.Close"    "JD.Close"    "RMO.Close"   
[6,] "NKLA.Volume"   "MPNGF.Volume"   "MSFT.Volume"   "JD.Volume"   "RMO.Volume"  
[7,] "NKLA.Adjusted" "MPNGF.Adjusted" "MSFT.Adjusted" "JD.Adjusted" "RMO.Adjusted"

The same code adjusted to fit in a specific environment:

library(quantmod)

symbolUpdates.env <- new.env()

# Load ticker data from 2020-01-01 till 2021-02-02 to symbolUpdates.env
t2 <- c("NKLA", "MPNGF", "RMO", "JD", "MSFT")
getSymbols.yahoo(t2, auto.assign = TRUE, env = symbolUpdates.env, 
                 from = "2020-01-01", to = "2021-02-02")

# Close all Internet connections as a precaution
closeAllConnections()

# Find xts objects in symbolUpdates.env
x2 <- names(which(unlist(eapply(symbolUpdates.env, is.xts))))

# Convert xts to data.frame that are in symbolUpdates.env
for (i2 in seq_along(x2)) {
  assign(envir = symbolUpdates.env, x2[i], fortify.zoo(get(x2[i2])))
}

# The query in symbolUpdates.env
sapply(mget(x2, envir = symbolUpdates.env), names)

# The rendering from symbolUpdates.env
     RMO            NKLA            JD            MSFT            MPNGF           
[1,] "Index"        "Index"         "Index"       "Index"         "Index"         
[2,] "Index"        "Index"         "Index"       "Index"         "Index"         
[3,] "RMO.Open"     "NKLA.Open"     "JD.Open"     "MSFT.Open"     "MPNGF.Open"    
[4,] "RMO.High"     "NKLA.High"     "JD.High"     "MSFT.High"     "MPNGF.High"    
[5,] "RMO.Low"      "NKLA.Low"      "JD.Low"      "MSFT.Low"      "MPNGF.Low"     
[6,] "RMO.Close"    "NKLA.Close"    "JD.Close"    "MSFT.Close"    "MPNGF.Close"   
[7,] "RMO.Volume"   "NKLA.Volume"   "JD.Volume"   "MSFT.Volume"   "MPNGF.Volume"  
[8,] "RMO.Adjusted" "NKLA.Adjusted" "JD.Adjusted" "MSFT.Adjusted" "MPNGF.Adjusted"

My questions:

Are there any mistakes in the code that an additional Index column appears in symbolUpdates.env?
If yes, what are they?
What could correct the issue?
Also, the order of the tickers is not respected in symbolUpdates.env, why?

Thanks in advance.

Systems used:

R version: 4.1.1 (2021-08-10)
RStudio version: 1.4.1717
OS: macOS Catalina version 10.15.7 and macOS Big Sur version 11.6

Wouldn't it be much much easier and cleaner to just use (named)lists instead of `assign` and environments ? — dario, Oct 08 '21 at 11:33
One error in your second version is in the `get()` call: `get(x2[i]))` defaults to the global environment, not `symbolUpdates.env`. — user2554330, Oct 08 '21 at 12:17
@user2554330, thanks for your comment. I corrected the mistake in the post. I did it in ```R``` too, but the problem persists: I still have 2 ```Index``` columns. — pdeli, Oct 08 '21 at 12:52
@dario. It is possible, however I have no idea how to do that. Would you mind showing how? — pdeli, Oct 08 '21 at 12:53
That is how we assign a value to named list in R: `new_list <- list(); new_list[["new_element"]] <- "new_element_value"`. — dario, Oct 08 '21 at 13:04
If you have further questions I'd suggest crating a [minimal reproducible example](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example/5963610#5963610). For example, if the question does not depend on the quantmod package remove it (I for one dont have that installed, so I won't be able to help) — dario, Oct 08 '21 at 13:06
@pdeli, you still have an error in the `get()`. Specifically, if you want to get something from a particular environment, you need to specify the `envir` argument. — user2554330, Oct 08 '21 at 13:56
@user2554330, thank you!M That did it. So ```for (i2 in seq_along(x2)) { assign(envir = symbolUpdates.env, x2[i2], fortify.zoo(get(x2[i]))) }``` should have been ```for (i2 in seq_along(x2)) { assign(envir = symbolUpdates.env, x2[i2], fortify.zoo(get(x2[i2], envir = symbolUpdates.env))) }```. Two errors in the line… — pdeli, Oct 08 '21 at 17:03

score 0 · Answer 1 · answered Oct 09 '21 at 04:34

The order can be arranged with match and for me I don't get any additional Index column.

lapply(mget(x2, envir = symbolUpdates.env), names)
data <- mget(x2[match(t2, x2)], envir = symbolUpdates.env)
sapply(data, names)

#      NKLA            MPNGF            RMO            JD            MSFT           
#[1,] "NKLA.Open"     "MPNGF.Open"     "RMO.Open"     "JD.Open"     "MSFT.Open"    
#[2,] "NKLA.High"     "MPNGF.High"     "RMO.High"     "JD.High"     "MSFT.High"    
#[3,] "NKLA.Low"      "MPNGF.Low"      "RMO.Low"      "JD.Low"      "MSFT.Low"     
#[4,] "NKLA.Close"    "MPNGF.Close"    "RMO.Close"    "JD.Close"    "MSFT.Close"   
#[5,] "NKLA.Volume"   "MPNGF.Volume"   "RMO.Volume"   "JD.Volume"   "MSFT.Volume"  
#[6,] "NKLA.Adjusted" "MPNGF.Adjusted" "RMO.Adjusted" "JD.Adjusted" "MSFT.Adjusted"

What I think is needed (not sure if this is true/necessary) is that the order of the content of ```symbolUpdates.env``` match the order of the content of ```.GlobalEnv``` (i.e., ```symbolUpdates.env``` content order = ```RMO NKLA JD MSFT MPNGF``` , whereas ```.GlobalEnv``` content order = ```MSFT MPNGF NKLA JD RMO```). — pdeli, Oct 09 '21 at 16:58
Nevertheless, ```identical(names(x), names(x2))``` renders ```TRUE```, so I guess ```data.frame``` objects order in different ```environment```s should not be a problem to append one to the other. Am I right? However, ```identical(x, x2)``` renders ```FALSE```… — pdeli, Oct 09 '21 at 17:20

pdeli · Answer 2 · 2021-10-09T16:34:07.440

@user2554330 pointed to the coding error and a piece missing. Correcting the first and adding the second did the trick.

Here is how the original code of the last for loop looked like:

for (i2 in seq_along(x2)) {
  assign(envir = symbolUpdates.env, x2[i], fortify.zoo(get(x2[i])))
}

Here is how it should have looked like (i2 instead of i and adding envir = symbolUpdates.env at the end:

for (i2 in seq_along(x2)) {
  assign(envir = symbolUpdates.env, x2[i2], fortify.zoo(get(x2[i2], 
                                           envir = symbolUpdates.env)))
}

Same code in different environments render different results in R

2 Answers2