How do I calculate percentiles/IQR with julia?

New to Julia
1

Hello,
I would like to find the percentile and interquartile range of a column of a dataframe.
I have loaded both Statistics and StatsBase, but when I try iqr(N[:column]) I get:

ERROR: MethodError: no method matching iqr(::Array{Union{Missing, Int64},1})
Closest candidates are:
  iqr(::AbstractArray{#s77,N} where N where #s77<:Real) at ~/.julia/packages/StatsBase/XePnn/src/scalarstats.jl:288
Stacktrace:
 [1] top-level scope at none:0

and similarly,

percentile(N[:column], 25)
ERROR: MethodError: no method matching percentile(::Array{Union{Missing, Int64},1}, ::Int64)
Closest candidates are:
  percentile(::AbstractArray{#s35,N} where N where #s35<:Real, ::Any) at ~/.julia/packages/StatsBase/XePnn/src/scalarstats.jl:174
Stacktrace:
 [1] top-level scope at none:0

Using the help function of REPL I got:

help?> iqr
search: iqr isqrt PartialQuickSort

  iqr(v)

  Compute the interquartile range (IQR) of an array, i.e. the 75th percentile minus the 25th percentile.
help?> percentile
search: percentile

  percentile(v, p)

  Return the pth percentile of a real-valued array v, i.e. quantile(x, p / 100).

Could these stats be computed on a dataframe? What am I getting wrong?
Thank you

2

Remove the missing values, eg collect(skipmissing(N[:column))).

3

There are no missing values:

julia> unique(df[:column]))
8-element Array{Union{Missing, Int64},1}:
 6064657
 6064658
 6064659
 6064660
 6064662
 6064666
 6064667
 6064668

and

julia> println(df[:column])
Union{Missing, Int64}[6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064657, 6064658, 6064659, 6064659, 6064660, 6064662, 6064666, 6064667, 6064668]
4

In that case, my suggestion would only changes the element type. This is needed because of

https://github.com/JuliaStats/StatsBase.jl/issues/449

5

Here’s an example:

D1 = DataFrame(a=randn(100))
allowmissing!(D1)
iqr(D1.a) # error
disallowmissing!(D1)
iqr(D1.a) # 1.517
6

That is nice, but I get the error MethodError: no method matching iterate(::DataFrame). Is it part of a specific package? REPL’s help only reports disallowmissing(x::AbstractArray)... and there was not much stuff on internet.

7

disallowmissing! is in the DataFrames package
Try either:

?disallowmissing!

methods(disallowmissing!)
8

That is weird then, since I loaded DataFrames already…