CODEWITHSUNDEEP
rfunctionuser-defined-functions
pachadotdev
pachadotdev

Reputation: 3775

How to use double paste in assign?

I'm working with different countries datasets downloaded from OEC's API.

I'm dividing my work in two steps:

  1. Getting and cleaning a certain dataset. OK
  2. Create a function to download any dataset. I'm stuck in this part

For the first part I did run

chile2013_df <- as.data.frame(fromJSON("http://atlas.media.mit.edu/hs92/export/2013/chl/all/show/"))
keep <- c("data.hs92_id", "data.import_val", "data.export_val")
chile2013_df <- chile2013_df[keep]

and that is ok and works, but if I want to create a function, let's say

country_data <- function(COUNTRYCODE, YEAR) {
assign(paste(COUNTRYCODE, YEAR, "_raw", sep=""), as.data.frame(fromJSON(paste("http://atlas.media.mit.edu/hs92/export/",YEAR,"/",COUNTRYCODE,"/all/show/", sep=""))), envir = globalenv());
assign(c("keep"), c("data.hs92_id", "data.import_val", "data.export_val");
assign(paste(COUNTRYCODE, YEAR, "_clean", sep=""), paste(COUNTRYCODE, YEAR, "_raw[keep]", sep=""));
envir = globalenv())
}

Then if I run

country_data("per","2010")

The raw file will be perfect but then per2010_clean will look as a text "per2010_raw[keep]"

How can I make per2010_raw[keep] effective? many thanks in advance.

Upvotes: 1

Views: 68

Answers (2)

chinsoon12
chinsoon12

Reputation: 25223

try this:

keep <- c("data.hs92_id", "data.import_val", "data.export_val")
country_data <- function(COUNTRYCODE, YEAR) {
    weblink <- paste0("http://atlas.media.mit.edu/hs92/export/",YEAR,"/",COUNTRYCODE,"/all/show/")
    rawdat <- as.data.frame(fromJSON(file=weblink))
    cleandat <- rawdat[keep]
    return(list(raw=rawdat, clean=cleandat))
} #end country_data

Then you can mapply for your pairs of (country, year). Ideally, dont use the country and year in your variable names as it can grow pretty unwieldy and unscalable after a while

#pulling data
alldat <- mapply(function(x, y) list(country_data(x, y)), c("per","chl"), c(2010, 2013))
names(alldat) <- paste0(c("per","chl"), c(2010, 2013))

#accessing data
alldat[["chl2013"]]$raw
alldat[["chl2013"]]$clean

Upvotes: 3

pachadotdev
pachadotdev

Reputation: 3775

Your approach is useful, thanks. Changing a little of your code I'm getting something closer to what I wanted.

keep <- c("data.hs92_id", "data.import_val", "data.export_val")
country_data <- function(COUNTRYCODE, YEAR) {
weblink <- paste0("http://atlas.media.mit.edu/hs92/export/",YEAR,"/",COUNTRYCODE,"/all/show/")
clean <<- as.data.frame(fromJSON(weblink))[keep]
}
country_data("per", 2010)

Upvotes: -1

Related Questions