Josh Day

Related: https://github.com/JuliaComputing/TextParse.jl/pull/131

For *distributed* tables, `getindex` (which gets called when you write `tbl[2]`) is not supported. I think we do need a better way of taking a glance at a distributed table.

Do you have an example we can reproduce?

Does this work for you? ```julia using JuliaDB, CSV table(CSV.File(filepath)) ```

You want to make use of the `colparsers` keyword arg for `loadtable`. Try ```julia loadtable(path, colparsers=Dict(:datecol => dateformat"MM/DD/YYYY")) ``` or similar (I may not have gotten that exactly right). https://juliacomputing.github.io/TextParse.jl/dev/#Reading-CSV-1

That's correct. Files aren't split into multiple chunks. Each chunk is at least one file.

`eltype(table)` is type stable: ```julia julia> @inferred eltype(table((x=1:5,y=1:5))) NamedTuple{(:x, :y),Tuple{Int64,Int64}} julia> @inferred eltype(table(1:5, 1:5)) Tuple{Int64,Int64} ``` That being said, I'm 100% on board for forcing the eltype to be a...

The `partitionplot` code needs some cleanup (see my stale PR https://github.com/JuliaComputing/JuliaDB.jl/pull/190). I haven't had time to look at it lately, but I may actually have the time today. Note that...

I don't think it would be too messy since it's only for Base.show methods. This is helpful when using the REPL, not needing to go back and forth between plotting...

1. I'm okay dropping any `_fun` function, but there are some micro-optimizations with things like `value_deriv` that can help you avoid calculating the same thing twice. 2. I think starting...