Plots borking on SentinelArrays produced by CSV.read

[digital_carver]

I came across this post

Disable SentinelArrays for CSV.read

This will be fixed soon. In general if you want to avoid SentinelVector (which often makes sense from my experience) set CSV.read to use a single thread when reading the data in (it will then be a bit slower but usually acceptably fast).

which recommends reading the CSV single-threaded. That works:

julia> wwdat1 = CSV.read("NWSSSarsCov2WastewaterConc.csv",DataFrame; ntasks = 1)
335417×4 DataFrame
...

julia> wwdat1.key_plot_id |> typeof
       # now it's a normal Vector, not a SentinelArray type
Vector{String} (alias for Array{String, 1})

julia> scatter(wwdat1.date,wwdat1.pcr_conc_smoothed) ## works.

The same thread also suggests collecting the column before plotting, as another way of managing this, and that works too:

julia> scatter(collect(wwdat.date), collect(wwdat.pcr_conc_smoothed))  # works

But reading the CSV with ntasks = 1 is a fixed, one-time (per session) cost, so that’s probably better than collecting for every plot call.