decimal-parameter for reading CSV-files?

[roland-KA]

Is there a decimal-parameter (as in CSV.jl) to specify if the fractional part of decimal numbers is separated by '.' (like 3.14) or by ',' (like 3,14) when reading CSV files?

[drizk1]

I will check but I'm not sure presently. Do you have a dataset you need to read w that format?

[roland-KA]

In Germany (and in many other European countries) the comma-separated format is standard. So it's an every-day use-case.

[drizk1]

Ok. If you have an easily accessible small dataset, please pass it along so I can begin testing with it as I figure out what csv.jl offers and how to make tidierfiles support it. If not no worries I'll find one to use

[roland-KA]

CommaExample.csv

This is an example with three columns A, B and C, where B and C are floating-point numbers using a comma-separator. The column-separator in these files is a semicolon (so that there isn't a confusion with the comma-separator within the floating-point numbers).

[drizk1]

ok, so there is an argument in csv.jl to choose what the decimal is.

read_csv, read_delim and read_tsv do not yet support that additional argument. To keep things consistent with the readr R package, I wil likely make a read_csv2 function that uses "','" as decimal and "'.'" as grouping mark

and then add options to read_delim to give users more control if needed.

[drizk1]

with v.1.2 which i should be on the registry in 20 mins or so
there are two ways to do it

• read_csv2 defaults to delim = ';' and decimal = ','
• for more control you can use read_delim as well

julia> read_csv2("path/to/CommaExample.csv")
10×3 DataFrame
 Row │ A        B        C       
     │ String1  Float64  Float64 
─────┼───────────────────────────
   1 │ j           90.0     89.0
   2 │ t           52.0      7.0
   3 │ z           47.0     85.0
   4 │ r           72.0     67.0
   5 │ j           50.0     85.0
   6 │ x            5.0     29.0
   7 │ a           84.0      7.0
   8 │ t           38.0     28.0
   9 │ w           96.0     42.0
  10 │ a           12.0     62.0

julia> read_delim("path/to/CommaExample.csv", decimal = ',', delim = ';')
10×3 DataFrame
 Row │ A        B        C       
     │ String1  Float64  Float64 
─────┼───────────────────────────
   1 │ j           90.0     89.0
   2 │ t           52.0      7.0
   3 │ z           47.0     85.0
   4 │ r           72.0     67.0
   5 │ j           50.0     85.0
   6 │ x            5.0     29.0
   7 │ a           84.0      7.0
   8 │ t           38.0     28.0
   9 │ w           96.0     42.0
  10 │ a           12.0     62.0

please reopen/let us know if you run into issues / want additional features

[roland-KA]

Wow, that went fast! Thank you very much 😊👍