AggregateBy

NOTE: AggregateBy is not registered. For a registered package with similar functionality, see https://github.com/JuliaData/SplitApplyCombine.jl.

AggregateBy supports simple aggregation operations on iterable containers. Functions that might (in other languages) be called countby and groupby are instead composed of two pieces, (1) an aggregation command and (2) a By(fkey, fval) elementwise operator that generates key, value pairs for aggregation. Some examples may help explain the general concept.

These examples assume that you've executed using AggregateBy in the current session.

To count all the letters in a string, ignoring case, use

julia> count(By(lowercase), "HelLo")
Dict{Char, Int64} with 4 entries:
  'h' => 1
  'l' => 2
  'e' => 1
  'o' => 1

To collect similar items, use collect:

julia> collect(By(isodd), 1:11)
Dict{Bool, Vector{Int64}} with 2 entries:
  0 => [2, 4, 6, 8, 10]
  1 => [1, 3, 5, 7, 9, 11]

If you have a matrix temperaturedata where the first column holds a DateTime (from the Dates library) and the last column holds the temperature, then the hourly maximum temperature can be computed as

julia> maximum(By(hour ∘ first, last), eachrow(temperaturedata))
Dict{Int64, Float64} with 8 entries:
  0  => 283.019
  15 => 287.615
  21 => 289.044
  6  => 278.94
  18 => 288.265
  9  => 279.512
  12 => 280.599
  3  => 277.989

See the documentation for further details and more options.