Read csv file with too many commas

Question

Maybe it's easy, but I have a csv file with a lot of commas and R doesn't read it correctly, it puts all data in the first column and doesn't present it as a table.

Do you know how I can make to read the file correctly as a classic csv file?

You can download the file here from world bank

Welcome to Stackoverflow. Please have a look at [how to make a good R reproducible example](https://stackoverflow.com/q/5963269/3250126). Please share the code and parts of the data. — loki, Jul 21 '17 at 08:37

Rui Barradas · Answer 1 · 2017-07-21T10:30:02.127

Well, it takes some time to clean the file, fortunately I have the time.

gdp2012 <- read.csv("getdata_data_GDP.csv", stringsAsFactors = FALSE)
cnames <- gdp2012[3, ]
names(cnames) <- NULL
cnames[1] <- "Abbrev"
cnames[5] <- "Millions.USD"
names(gdp2012) <- cnames
names(gdp2012)
[1] "Abbrev"       "Ranking"      "NA"           "Economy"          "Millions.USD"
[6] ""             "NA"           "NA"           "NA"           "NA"

gdp2012 <- gdp2012[, -grep("NA", names(gdp2012))]
gdp2012 <- gdp2012[, -ncol(gdp2012)]
gdp2012 <- gdp2012[-c(1:4, 237:nrow(gdp2012)), ]

dim(gdp2012)
[1] 232   4

str(gdp2012)
'data.frame':   232 obs. of  4 variables:
 $ Abbrev      : chr  "USA" "CHN" "JPN" "DEU" ...
 $ Ranking     : chr  "1" "2" "3" "4" ...
 $ Economy     : chr  "United States" "China" "Japan" "Germany" ...
 $ Millions.USD: chr  " 16,244,600 " " 8,227,103 " " 5,959,718 " "  3,428,131 " ...

gdp2012[[4]] <- as.numeric(gsub(",", "", gdp2012[[4]]))
Warning message:
NAs introduced by coercion

gdp2012[[2]] <- as.numeric(gdp2012[[2]])

head(gdp2012)
   Abbrev Ranking        Economy Millions.USD
5     USA       1  United States     16244600
6     CHN       2          China      8227103
7     JPN       3          Japan      5959718
8     DEU       4        Germany      3428131
9     FRA       5         France      2612878
10    GBR       6 United Kingdom      2471784

If you want the row numbers to start at 1, just do

rownames(gdp2012) <- NULL

Thank you very much !! I'm starting in r, and this sort of file is not easy for me... It's yet friday for me, and worktime... haha — Aurélien, Jul 21 '17 at 10:10

score 0 · Answer 2 · answered Jul 21 '17 at 08:54

0

Delete row 1,-3,238-241 in csv itself

use rio library. data=import("filename").

There are lots of blank rows in data. so you can use data[rowSums(is.na(data)) == 0,]

answered Jul 21 '17 at 08:54

PritamJ

337
4
10

score 0 · Answer 3 · answered Jul 21 '17 at 10:37

Using data.table, fread function, which has many import arguments and also faster:

library(data.table)

res <- fread("myFile.csv",
             sep = ",", # separator is comma
             skip = 5,  # skip first 5 rows
             select = c(1, 2, 4, 5), # select columns by index
             na.strings = c("", ".."), # convert blanks to NA
             # set column names
             col.names = c("Country", "Ranking", "Economy", "USD_Mln"))

# remove blank rows
res <- res[ !is.na(Country), ]

# convert character numbers to numbers
res[ , USD_Mln := as.numeric(gsub(",", "", USD_Mln))]

head(res)
#    Country Ranking        Economy  USD_Mln
# 1:     USA       1  United States 16244600
# 2:     CHN       2          China  8227103
# 3:     JPN       3          Japan  5959718
# 4:     DEU       4        Germany  3428131
# 5:     FRA       5         France  2612878
# 6:     GBR       6 United Kingdom  2471784

Read csv file with too many commas

3 Answers3