# When colwise generate a couple of arrays instead of one array with many elements
**URL:** https://discourse.julialang.org/t/when-colwise-generate-a-couple-of-arrays-instead-of-one-array-with-many-elements/663
**Category:** Data
**Tags:** question
**Created:** [November 30, 2016, 11:59pm UTC](https://discourse.julialang.org/t/when-colwise-generate-a-couple-of-arrays-instead-of-one-array-with-many-elements/663 "2016-11-30T23:59:12Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@Hongwei](https://discourse.julialang.org/u/Hongwei)
#### Post date: [November 30, 2016, 11:59pm UTC](https://discourse.julialang.org/t/when-colwise-generate-a-couple-of-arrays-instead-of-one-array-with-many-elements/663/1 "2016-11-30T23:59:12Z")
Hi guys,
cf = colwise(mean, test) #test is a two columns data frame
I got:
dump(cf)
Array{Any}((2,))
1: Array{Float64}((1,)) [0.056307]
2: Array{Float64}((1,)) [0.0330923]
So I have to use [float(cf[i][1]) for i in 1:size(cf)[1]] to get this:
2-element Array{Float64,1}:
0.056307
0.0330923
Is there an easy way to get Array{Float64,1} instead of Array{Any}((2,))?
Thanks,
Hongwei
---
### Author:  [@Hongwei](https://discourse.julialang.org/u/Hongwei)
#### Post date: [December 1, 2016, 3:10am UTC](https://discourse.julialang.org/t/when-colwise-generate-a-couple-of-arrays-instead-of-one-array-with-many-elements/663/2 "2016-12-01T03:10:36Z")
Just find squeeze(mean(Array(test), 1), 1) could do the same thing.
---
### Author:  [@nalimilan](https://discourse.julialang.org/u/nalimilan)
#### Post date: [December 2, 2016, 2:19pm UTC](https://discourse.julialang.org/t/when-colwise-generate-a-couple-of-arrays-instead-of-one-array-with-many-elements/663/3 "2016-12-02T14:19:53Z")
Yes, that’s [a known issue](https://github.com/JuliaStats/DataFrames.jl/issues/485). For now your solution is the best workaround.