Column names mis-match when loading certain columns

[Ossifragus]

The relative order of the loaded columns seem to be always the same of the relative order the whole data (Is the order specified in the metadata?). For example with a data consisting two columns "col1" and "col2", the code JDF.load(data; cols = ["col1", "col2"]) loads the data correctly, but with JDF.load(data; cols = ["col2", "col1"]) the data columns and their names would not match.

Please see the following example:

using RDatasets, JDF, DataFrames

JDF.save("iris.jdf", dataset("datasets", "iris"))
a = DataFrame(JDF.load("iris.jdf"))

c1 = ["Species", "PetalWidth"]
a1 = DataFrame(JDF.load("iris.jdf"; cols = c1)) # column mis-match

julia> first(a1, 5)
5×2 DataFrame
 Row │ Species  PetalWidth 
     │ Float64  Cat…       
─────┼─────────────────────
   1 │     0.2  setosa
   2 │     0.2  setosa
   3 │     0.2  setosa
   4 │     0.2  setosa
   5 │     0.2  setosa


c2 = ["PetalWidth", "Species"]
a2 = DataFrame(JDF.load("iris.jdf"; cols = c2)) # column match

julia> first(a2, 5)
5×2 DataFrame
 Row │ PetalWidth  Species 
     │ Float64     Cat…    
─────┼─────────────────────
   1 │        0.2  setosa
   2 │        0.2  setosa
   3 │        0.2  setosa
   4 │        0.2  setosa
   5 │        0.2  setosa

[xiaodaigh]

columns are loaded as specified in the order by cols. This is the intended behaviour though

[xiaodaigh]

you very right. i missed that detail.

It's been fixed with the latest version of v0.5

[Ossifragus]

Great! Thank you so much!

[4db83]

Hi, I still have the same problem with this. That is, the cols ordering is not satisfied. See code snipped below (Julia v1.7):

(@v1.7) pkg> status JDF
Status C:\Users\ibm\.julia\environments\v1.7\Project.toml
[babc3d20] JDF v0.5.0

julia> first(DataFrame(JDF.load("gkx-msf-merged.jdf", cols = [:date, :permno])),5)
5×2 DataFrame
Row │ permno date
│ Int64 Int64
─────┼──────────────────
1 │ 10000 19860228
2 │ 10000 19860331
3 │ 10000 19860430
4 │ 10000 19860530
5 │ 10000 19860630

julia> first(DataFrame(JDF.load("gkx-msf-merged.jdf", cols = [:permno, :date])),5)
5×2 DataFrame
Row │ permno date
│ Int64 Int64
─────┼──────────────────
1 │ 10000 19860228
2 │ 10000 19860331
3 │ 10000 19860430
4 │ 10000 19860530
5 │ 10000 19860630

julia>

[xiaodaigh]

I run the load in parallel so it's hard to force the column order. let me see

[xiaodaigh]

fixed by #81

[4db83]

Thank you From: evalparse ***@***.***> Sent: Wednesday, February 9, 2022 11:58 To: xiaodaigh/JDF.jl ***@***.***> Cc: 4db83 ***@***.***>; Comment ***@***.***> Subject: Re: [xiaodaigh/JDF.jl] Column names mis-match when loading certain columns (Discussion #78) fixed by #81 <#81> — Reply to this email directly, view it on GitHub <#78 (reply in thread)> , or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUGJQHWTOENBZJWUKRYNDILU2JCCXANCNFSM5M4ZHS5A> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub> . You are receiving this because you commented.Message ID: ***@***.***>