Dataframe columns type is `Any` even though it's initialized with all `Float64`

New to Julia
,
1

I don’t have much more to add, I even rerun the code but initializing the columns with the type as a string, and still:

image
image218×321 2.77 KB

The two columns before them are fine, but no matter what i do these have type Any, and that’s making my code throw errors later on

2

How are you initializing?


julia> A = Array{Float64}(undef, 4, 4)
4×4 Matrix{Float64}:
 0.0  0.0  0.0  0.0
 0.0  0.0  0.0  0.0
 0.0  0.0  0.0  0.0
 0.0  0.0  0.0  0.0

julia> df = DataFrame(A, :auto)
4×4 DataFrame
 Row │ x1       x2       x3       x4      
     │ Float64  Float64  Float64  Float64 
─────┼────────────────────────────────────
   1 │     0.0      0.0      0.0      0.0
   2 │     0.0      0.0      0.0      0.0
   3 │     0.0      0.0      0.0      0.0
   4 │     0.0      0.0      0.0      0.0
3

I used some arrays of floats i constructed before.

Oh, i found the problem. I was initializing the vectors as [], with no values inside them and no type, so they stayed as Vector{Any}s.
Changing the declarations to Vector{Float64}() solved the problem.

2 Likes
4

You can use also Float64[] (and more generally T[]) to initialize an empty vector or Float64…

2 Likes