From 0d1fe72d538d535c9b3b793fa868b93119ed0688 Mon Sep 17 00:00:00 2001
From: Milan Bouchet-Valat <nalimilan@club.fr>
Date: Sat, 2 Mar 2019 11:45:47 +0100
Subject: [PATCH] Skip missing values in describe (#470)

---
 src/scalarstats.jl | 30 ++++++++++++++----------------
 test/misc.jl       |  7 ++++++-
 2 files changed, 20 insertions(+), 17 deletions(-)

diff --git a/src/scalarstats.jl b/src/scalarstats.jl
index a885e6493115e..5a53b34f225a2 100644
--- a/src/scalarstats.jl
+++ b/src/scalarstats.jl
@@ -562,17 +562,19 @@ median, 75th percentile, and maxmimum.
 function summarystats(a::AbstractArray{T}) where T<:Union{Real,Missing}
     # `mean` doesn't fail on empty input but rather returns `NaN`, so we can use the
     # return type to populate the `SummaryStats` structure.
-    m = mean(a)
+    s = T >: Missing ? collect(skipmissing(a)) : a
+    m = mean(s)
     R = typeof(m)
     n = length(a)
-    qs = if n == 0
+    ns = length(s)
+    qs = if m == 0 || n == 0
         R[NaN, NaN, NaN, NaN, NaN]
-    elseif ismissing(m)
-        [missing, missing, missing, missing, missing]
+    elseif T >: Missing
+        quantile!(s, [0.00, 0.25, 0.50, 0.75, 1.00])
     else
-        quantile(a, [0.00, 0.25, 0.50, 0.75, 1.00])
+        quantile(s, [0.00, 0.25, 0.50, 0.75, 1.00])
     end
-    SummaryStats{R}(m, qs..., n, count(ismissing, a))
+    SummaryStats{R}(m, qs..., n, n - ns)
 end
 
 function Base.show(io::IO, ss::SummaryStats)
@@ -580,16 +582,12 @@ function Base.show(io::IO, ss::SummaryStats)
     @printf(io, "Length:         %i\n", ss.nobs)
     ss.nobs > 0 || return
     @printf(io, "Missing Count:  %i\n", ss.nmiss)
-    if ss.nmiss > 0
-        println(io, "(All summary stats are missing)")
-    else
-        @printf(io, "Mean:           %.6f\n", ss.mean)
-        @printf(io, "Minimum:        %.6f\n", ss.min)
-        @printf(io, "1st Quartile:   %.6f\n", ss.q25)
-        @printf(io, "Median:         %.6f\n", ss.median)
-        @printf(io, "3rd Quartile:   %.6f\n", ss.q75)
-        @printf(io, "Maximum:        %.6f\n", ss.max)
-    end
+    @printf(io, "Mean:           %.6f\n", ss.mean)
+    @printf(io, "Minimum:        %.6f\n", ss.min)
+    @printf(io, "1st Quartile:   %.6f\n", ss.q25)
+    @printf(io, "Median:         %.6f\n", ss.median)
+    @printf(io, "3rd Quartile:   %.6f\n", ss.q75)
+    @printf(io, "Maximum:        %.6f\n", ss.max)
 end
 
 
diff --git a/test/misc.jl b/test/misc.jl
index 7c09b32b7f542..6773d2c2a58d0 100644
--- a/test/misc.jl
+++ b/test/misc.jl
@@ -56,7 +56,12 @@ describe(io, Union{Float32,Missing}[1.0, 4.5, missing, missing, 33.1])
                            Summary Stats:
                            Length:         5
                            Missing Count:  2
-                           (All summary stats are missing)
+                           Mean:           12.866666
+                           Minimum:        1.000000
+                           1st Quartile:   2.750000
+                           Median:         4.500000
+                           3rd Quartile:   18.799999
+                           Maximum:        33.099998
                            Type:           $(Union{Float32,Missing})
                            """