You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm using metadata for the first time today (I usually use Stata and R for data cleaning) and I found it helpful to define the following functions to make it easier to see metadata
function showcolmetadata(df, col::Union{AbstractString, Symbol}, label::AbstractString)
colmetadata(df, col, label)
end
function showcolmetadata(df, cols::AbstractVector{<:Union{AbstractString, Symbol}}, label::AbstractString)
labvals = [colmetadata(df, col, label) for col in cols]
nt = (; :variable => cols, Symbol(label) => labvals)
pretty_table(nt; nosubheader = true, crop = :horizontal, alignment = [:l, :l])
end
function showcolmetadata(df, col::Union{AbstractString, Symbol}, labels::AbstractVector{<:AbstractString})
labvals = [colmetadata(df, col, label) for label in labels]
nt(; :label => labels, :value => labvals)
pretty_table(nt; nosubheader = true, tile = "colmetadata for $col", crop = :horizontal, alignment = [:l, :l])
end
function showcolmetadata(df, cols::Colon, label::AbstractString)
showcolmetadata(df, names(df), label)
end
function showcolmetadata(df, col::Union{AbstractString, Symbol}, labels::Colon)
d = colmetadata(df, col)
labnames = names(d)
showcolmetadata(df, col, labnames)
end
They would be used like following: You can see that the column names are pretty uninformative in the Penn World Tables, exactly the kind of situation metadata is great for.
julia> showcolmetadata(pwt[:, 1:10], :, "label")
┌───────────────┬──────────────────────────────────────────────────────────
│ variable │ label ⋯
├───────────────┼──────────────────────────────────────────────────────────
│ countrycode │ 3-letter ISO country code ⋯
│ country │ Country name ⋯
│ currency_unit │ Currency unit ⋯
│ year │ Year ⋯
│ rgdpe │ Expenditure-side real GDP at chained PPPs (in mil. 2017 ⋯
│ rgdpo │ Output-side real GDP at chained PPPs (in mil. 2017US$) ⋯
│ pop │ Population (in millions) ⋯
│ emp │ Number of persons engaged (in millions) ⋯
│ avh │ Average annual hours worked by persons engaged (source: ⋯
│ hc │ Human capital index, see note hc ⋯
└───────────────┴──────────────────────────────────────────────────────────
1 column omitted
The text was updated successfully, but these errors were encountered:
I'm using metadata for the first time today (I usually use Stata and R for data cleaning) and I found it helpful to define the following functions to make it easier to see metadata
They would be used like following: You can see that the column names are pretty uninformative in the Penn World Tables, exactly the kind of situation metadata is great for.
The text was updated successfully, but these errors were encountered: