Add rebinning hints (#100)

* Add rebinning hints

* Improve error message

* Bump version number

Project.toml

@@ -1,7 +1,7 @@
 name = "FHist"
 uuid = "68837c9b-b678-4cd5-9925-8a54edc8f695"
 authors = ["Moelf <proton[at]jling.dev>", "Nick Amin <amin.nj[at]gmail.com>"]
-version = "0.10.8"
+version = "0.10.9"
 
 [deps]
 BayesHistogram = "000d9b38-65fe-4c81-bdb9-69f01f102479"

src/hist1d.jl

@@ -309,7 +309,10 @@ Merges `n` consecutive bins into one.
 The returned histogram will have `nbins(h)/n` bins.
 """
 function rebin(h::Hist1D, n::Int=1)
-    @assert nbins(h) % n == 0
+    if nbins(h) % n != 0
+        rebin_values = join(sort(collect(valid_rebin_values(h))), ", ", " or ")
+        error("Invalid rebin value ($n) for a 1D histogram with $(nbins(h)) bins. It has to be a multiple of $(rebin_values)")
+    end
     p = x->Iterators.partition(x, n)
     counts = sum.(p(bincounts(h)))
     sumw2 = sum.(p(h.sumw2))

src/hist2d.jl

@@ -238,7 +238,10 @@ Merges `nx` (`ny`) consecutive bins into one along the x (y) axis by summing.
 """
 function rebin(h::Hist2D, nx::Int=1, ny::Int=nx)
     sx, sy = nbins(h)
-    @assert sx % nx == sy % ny == 0
+    if !(sx % nx == sy % ny == 0)
+        rebin_values_x, rebin_values_y = map(x->join(sort(collect(x)), ", ", " or "), valid_rebin_values(h))
+        error("Invalid rebin values (nx: $nx, ny: $ny) for a 2D histogram with $(nbins(h)) bins. They have to be a multple of nx: $(rebin_values_x) and ny: $(rebin_values_y)")
+    end
     p1d = (x,n)->Iterators.partition(x, n)
     p2d = x->(x[i:i+(nx-1),j:j+(ny-1)] for i=1:nx:sx, j=1:ny:sy)
     counts = sum.(p2d(bincounts(h)))
@@ -251,6 +254,7 @@ function rebin(h::Hist2D, nx::Int=1, ny::Int=nx)
 end
 rebin(nx::Int, ny::Int) = h::Hist2D -> rebin(h, nx, ny)
 
+
 """
     project(h::Hist2D, axis::Symbol=:x)
     project(axis::Symbol=:x) = h::Hist2D -> project(h, axis)

src/utils.jl

@@ -35,3 +35,33 @@ end
 function _is_uniform_bins(A::AbstractRange{T}) where T<:Real
     true
 end
+
+"""
+    valid_rebin_values(h::Union{Hist1D, Hist2D, Hist3D})
+
+Calculates the legal values for rebinning, essentially the prime factors of
+the number of bins. For a 1D histogram, a `Set` of numbers is return, for higher
+dimensional histograms a `Vector{Set}` for each dimension.
+"""
+valid_rebin_values(h::Hist1D) = _factor(nbins(h))
+valid_rebin_values(h::Union{Hist2D, Hist3D}) = [_factor(x) for x in nbins(h)]
+
+"""
+    function _factor(n::Integer)
+
+Helper function to calculate the prime factors of a given integer.
+"""
+function _factor(n::Integer)
+    factors = Set{Int}()
+    limit = n
+    factor = 2
+    while n > 1 & factor < limit
+        while n % factor == 0
+            n /= factor
+            push!(factors, factor)
+            limit = sqrt(n)
+        end
+        factor += 1
+    end
+    factors
+end

test/runtests.jl

@@ -419,13 +419,15 @@ end
     @test integral(h1) == integral(rebin(h1, 5))
     @test sum(h1.sumw2) == sum(rebin(h1, 5).sumw2)
     @test binedges(rebin(h1, 5)) == [0, 0.5, 1.0]
+    @test Set([2, 5]) == FHist.valid_rebin_values(h1)
 
     h2 = Hist1D(rand(10^2), [0.0, 0.1, 0.7, 0.9, 1.0])
     @test h2 == rebin(h2, 1)
     @test integral(h2) == integral(rebin(h2, 2))
     @test nentries(h2) == nentries(rebin(h2, 1))
     @test sum(h2.sumw2) == sum(rebin(h2, 2).sumw2)
     @test binedges(rebin(h2, 2)) == [0, 0.7, 1.0]
+    @test Set([2]) == FHist.valid_rebin_values(h2)
 
     @test rebin(h1, 2) == (h1 |> rebin(2))
 
@@ -434,17 +436,20 @@ end
     @test integral(h1) == integral(rebin(h1, 5))
     @test sum(h1.sumw2) == sum(rebin(h1, 5).sumw2)
     @test binedges(rebin(h1, 5)) == ([0, 0.5, 1.0], [0, 0.5, 1.0])
+    @test Set([2]) == FHist.valid_rebin_values(h2)
 
     bins = [0.0, 0.1, 0.7, 0.9, 1.0]
     h2 = Hist2D((rand(10^2),rand(10^2)), (bins,bins))
     @test h2 == rebin(h2, 1) == (h2 |> rebin(1, 1))
     @test integral(h2) == integral(rebin(h2, 2))
     @test sum(h2.sumw2) == sum(rebin(h2, 2).sumw2)
     @test binedges(rebin(h2, 2)) == ([0, 0.7, 1.0], [0, 0.7, 1.0])
+    @test [Set([2]), Set([2])] == FHist.valid_rebin_values(h2)
 
     h2 = Hist2D((rand(10^2),rand(10^2)), (0:0.1:1,0:0.5:1))
     @test nbins(rebin(h2, 10, 2)) == (1, 1)
-    @test_throws AssertionError rebin(h2, 2, 10)
+    @test_throws ErrorException rebin(h2, 2, 10)
+    @test [Set([5, 2]), Set([2])] == FHist.valid_rebin_values(h2)
 end
 
 @testset "Profile" begin