Add function to join community profiles (#58)

* add function to join community profiles

* add news, bump minor version

* add tests

NEWS.md

@@ -1,5 +1,9 @@
 # News for Microbiome.jl
 
+## Unreleased
+
+- Add `commjoin` function to merge the contents of multiple `CommunityProfile`s ([#58](https://github.com/BioJulia/Microbiome.jl/pull/58))
+
 ## v0.7.0 Major overhaul
 
 Lots of breaking changes in this one, including

Project.toml

@@ -4,7 +4,7 @@ keywords = ["microbiology", "microbiome", "biology"]
 license = "MIT"
 desc = "Functions and types for working with microbial community data"
 authors = ["@kescobo <[email protected]>"]
-version = "0.7.1"
+version = "0.7.2"
 
 [deps]
 AxisIndices = "f52c9ee2-1b1c-4fd8-8546-6350938c7f11"

docs/src/profiles.md

@@ -6,6 +6,7 @@ samples
 features
 samplenames
 featurenames
+commjoin
 relativeabundance
 relativeabundance!
 present

src/Microbiome.jl

@@ -36,7 +36,8 @@ export CommunityProfile,
        clades,
        profiletype,
        featuretotals,
-       sampletotals
+       sampletotals,
+       commjoin
 
 # Abundances
 export present,
@@ -73,5 +74,6 @@ include("ecobase.jl")
 include("samples_features.jl")
 include("profiles.jl")
 include("diversity.jl")
+include("comm_joins.jl")
 
 end  # module Microbiome

src/comm_joins.jl

@@ -0,0 +1,73 @@
+"""
+    commjoin(c1::CommunityProfile, comms::CommunityProfile...)
+
+Join multiple `CommunityProfile`s, creating a new `CommunityProfile`.
+For now, sample names cannot overlap in any of the input profiles.
+
+```jldoctest
+julia> mss = [MicrobiomeSample("sample\$i") for i in 1:15];
+
+julia> txs = [Taxon("taxon\$i") for i in 1:20];
+
+julia> cm1 = CommunityProfile(spzeros(10,5), txs[1:10], mss[1:5])
+CommunityProfile{Float64, Taxon, MicrobiomeSample} with 10 things in 5 places
+
+Thing names:
+taxon1, taxon2, taxon3...taxon9, taxon10
+
+Place names:
+sample1, sample2, sample3, sample4, sample5
+
+
+
+julia> cm2 = CommunityProfile(spzeros(10,5), txs[6:15], mss[6:10])
+CommunityProfile{Float64, Taxon, MicrobiomeSample} with 10 things in 5 places
+
+Thing names:
+taxon6, taxon7, taxon8...taxon14, taxon15
+
+Place names:
+sample6, sample7, sample8, sample9, sample10
+
+
+
+julia> cm3 = CommunityProfile(spzeros(10,5), txs[11:20], mss[11:15])
+CommunityProfile{Float64, Taxon, MicrobiomeSample} with 10 things in 5 places
+
+Thing names:
+taxon11, taxon12, taxon13...taxon19, taxon20
+
+Place names:
+sample11, sample12, sample13, sample14, sample15
+
+
+
+julia> commjoin(cm1, cm2, cm3)
+CommunityProfile{Float64, Taxon, MicrobiomeSample} with 20 things in 15 places
+
+Thing names:
+taxon1, taxon2, taxon3...taxon19, taxon20
+
+Place names:
+sample1, sample2, sample3...sample14, sample15
+```
+"""
+function commjoin(c1::CommunityProfile, comms::CommunityProfile...)
+    length(intersect(samplenames(c1), samplenames.(comms)...)) == 0 || error("Duplicate sample names detected: $(intersect(samplenames(c1), samplenames.(comms)...))")
+
+    all_samples = vcat(samples(c1), samples.(comms)...)
+    sample_dict = dictionary(zip(name.(all_samples), eachindex(all_samples)))
+    all_features = unique(vcat(features(c1), features.(comms)...))
+    feature_dict = dictionary(zip(name.(all_features), eachindex(all_features)))
+
+    mat = spzeros(length(all_features), length(all_samples))
+    for comm in (c1, comms...)
+        for sample in samplenames(comm)
+            for feature in featurenames(comm)
+                mat[feature_dict[feature], sample_dict[sample]] = comm[feature, sample]
+            end
+        end
+    end
+
+    return CommunityProfile(mat, all_features, all_samples)
+end

test/runtests.jl

@@ -125,6 +125,13 @@ end
         end
 
         @test_throws ArgumentError relativeabundance!(comm, kind=:invalid)
+
+        @test_throws ErrorException commjoin(comm, comm)
+        let c3 = commjoin(comm[:,1:2], comm[:, 3:4], comm[:, 5])
+            @test abundances(c3) == abundances(comm)
+            @test samples(c3) == samples(comm)
+            @test features(c3) == features(comm)
+        end
     end
 
     @testset "Indexing and Tables integration" begin