Major refactor to push generic methods to other packages

.travis.yml

@@ -4,8 +4,8 @@ os:
   - linux
   - osx
 julia:
-  - 1.0
   - 1.1
+  - 1.2
   - nightly
 notifications:
   email: false
@@ -24,19 +24,23 @@ matrix:
 #before_script: # homebrew for mac
 #  - if [ $TRAVIS_OS_NAME = osx ]; then brew install gcc; fi
 
-# # uncomment the following lines to override the default test script
-# script:
-#  - julia -e 'Pkg.clone(pwd()); Pkg.build("Microbiome"); Pkg.test("Microbiome"; coverage=true)'
+# comment the following lines to use the default test script
+script:
+  - julia -e 'using Pkg.Registry; Registry.add(Registry.RegistrySpec(url = "https://github.com/BioJulia/BioJuliaRegistry.git"))'
+  - julia -e 'using Pkg.Registry; Registry.add(Registry.RegistrySpec(url = "https://github.com/JuliaRegistries/General.git"))'
+  - julia -e 'using Pkg; Pkg.build(); Pkg.test("Microbiome", coverage = true)'
 after_success:
   - julia --project=coverage/ -e 'using Pkg; Pkg.instantiate()'
   - julia --project=coverage/ coverage/coverage.jl
 
 jobs:
   include:
     - stage: "Documentation"
-      julia: 1.0
+      julia: 1.1
       os: linux
       script:
+        - julia -e 'using Pkg.Registry; Registry.add(Registry.RegistrySpec(url = "https://github.com/BioJulia/BioJuliaRegistry.git"))'
+        - julia -e 'using Pkg.Registry; Registry.add(Registry.RegistrySpec(url = "https://github.com/JuliaRegistries/General.git"))'
         - julia --project=docs/ -e 'using Pkg; Pkg.instantiate();
                                     Pkg.develop(PackageSpec(path=pwd()))'
         - julia --project=docs/ docs/make.jl

NEWS.md

@@ -1,5 +1,13 @@
 # News for Microbiome.jl
 
+## v0.5.0
+
+Major Changes:  
+- Dropped Microbiome.jl-defined types and methods for DistanceMatrix. Use Distances.jl instead. [Requires this SpatialEcology PR](https://github.com/EcoJulia/SpatialEcology.jl/pull/36)
+- Dropped Microbiome.jl-defined types and methods for PCoA. Use MDS from MultivariateStats.jl instead. [Requires this MultivariateStats PR](https://github.com/JuliaStats/MultivariateStats.jl/pull/85)
+- Dropped Hclust leaf ordering. Added to Clustering.jl instead ([see Clustering.jl PR](https://github.com/JuliaStats/Clustering.jl/pull/170))
+
+
 ## v0.4.1
 
 Major Changes:  

Project.toml

@@ -3,26 +3,25 @@ uuid = "3bd8f0ae-a0f2-5238-a5af-e1b399a4940c"
 keywords = ["microbiology", "microbiome", "biology"]
 license = "MIT"
 desc = "Functions for working with microbial community data"
-authors = ["kescobo <[email protected]"]
-version = "0.4.1"
+authors = ["kescobo <[email protected]>"]
+version = "0.5.0"
 
 [deps]
-Clustering = "aaaa29a8-35af-508c-8bc3-b662a17a0fe5"
 DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
-Distances = "b4f34e82-e78d-54a5-968a-f98e89d6e8f7"
-LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
+MultivariateStats = "6f286f6a-111f-5878-ab1e-185364afe411"
 Reexport = "189a3867-3050-52da-a836-e630ba90ab69"
 SpatialEcology = "348f2d5d-71a3-5ad4-b565-8af070f99681"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 
 [compat]
-julia = "1.0, 1.1"
+julia = "1.1, 1.2"
 
 [extras]
 Coverage = "a2441757-f6aa-5fb2-8edb-039e3f45d037"
+Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [targets]
 coverage = ["Coverage"]
-test = ["Test"]
+test = ["Test", "Random"]

appveyor.yml

@@ -1,7 +1,7 @@
 environment:
   matrix:
-  - julia_version: 1.0
   - julia_version: 1.1
+  - julia_version: 1.2
   - julia_version: latest
 
 platform:

docs/Project.toml

@@ -1,5 +1,14 @@
 [deps]
+Clustering = "aaaa29a8-35af-508c-8bc3-b662a17a0fe5"
+DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
+Distances = "b4f34e82-e78d-54a5-968a-f98e89d6e8f7"
+Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
 Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
+Microbiome = "3bd8f0ae-a0f2-5238-a5af-e1b399a4940c"
+MicrobiomePlots = "bed85cf4-5cc8-5ac4-9eb3-09fbf92b2ce2"
+MultivariateStats = "6f286f6a-111f-5878-ab1e-185364afe411"
+SpatialEcology = "348f2d5d-71a3-5ad4-b565-8af070f99681"
+StatsPlots = "f3b207a7-027a-5e70-b257-86293d7955fd"
 
 [compat]
-Documenter = "~0.19"
+Documenter = "~0.22"

docs/make.jl

@@ -1,15 +1,16 @@
 using Documenter, Microbiome
 
 makedocs(
-    format = :html,
     sitename = "Microbiome.jl",
     pages = [
         "Home" => "index.md",
         "Microbial Abundances" => "abundances.md",
         "Distances & Dissimilarity" => "distances.md",
         "Contributing" => "contributing.md"
     ],
-    authors = "Kevin Bonham, PhD"
+    authors = "Kevin Bonham, PhD",
+    format = Documenter.HTML(
+        prettyurls = get(ENV, "CI", nothing) == "true")
 )
 
 deploydocs(

docs/src/abundances.md

@@ -6,7 +6,7 @@ names are also stored, and there's a convenience function if you want to convert
 a `DataFrame` to a `ComMatrix`, assuming the first column contains feature
 names:
 
-```@repl 1
+```@example 1
 using Microbiome
 using DataFrames
 
@@ -22,7 +22,7 @@ Forgive the clutter... ComMatricies name rows as species (which is true in this
 case, but need not be), and columns are "sites" rather than samples. That will
 be fixed eventually.
 
-```@repl 1
+```@example 1
 samplenames(abund)
 featurenames(abund)
 sampletotals(abund) # this is column sums
@@ -32,7 +32,7 @@ featuretotals(abund) # this is row sums
 If you want relative abundance, you can do `relativeabundance(abund)` or
 `relativeabundance!(abund)`:
 
-```@repl 1
+```@example 1
 relativeabundance!(abund)
 
 sampletotals(abund)
@@ -43,33 +43,61 @@ function automatically generates an `n+1` row for `other` containing the
 remaining features. Note - these doesn't modify in-place, so you've gotta
 reassign if you want to update:
 
-```@repl 1
+```@example 1
 abund2 = filterabund(abund, 1)
 
 featurenames(abund2)
 ```
 
 ## Plotting
 
-**NOTE: The following functions are not currently working - I've moved them to a new package to simplify dependencies. I'm leaving the docs for now as a reference - see `Microbiome.jl` versions 0.2.1 and below for working versions**
+Some convenience plotting types are available using
+[MicrobiomePlots](https://github.com/BioJulia/MicrobiomePlots.jl)
+and [StatsPlots](https://github.com/juliaplots/StatsPlots.jl)
+
+```@example 1
+ENV["GKSwstype"] = "100" # hide
+using StatsPlots
+using MicrobiomePlots
+using Distributions
+using Random # hide
+Random.seed!(1) # hide
+
+# add some high abundance bugs to be a bit more realistic
+function spikein(spikes, y, x)
+    m = rand(LogNormal(), y, x)
+    for s in spikes
+        m[s, :] = rand(LogNormal(3., 0.5), x)'
+    end
+    return m
+end
+
+# 100 species in 10 samples, with every 10th bug a bit more abundant
+bugs = spikein(1:10:100, 100, 10)
+
+abund = abundancetable(bugs,
+    ["sample_$x" for x in 1:10],
+    ["species_$x" for x in 1:100]);
 
-Some convenience plotting types are available using [`RecipesBase`](https://github.com/juliaplots/recipesbase.jl) and
-[StatPlots](https://github.com/juliaplots/StatPlots.jl)
+relativeabundance!(abund)
+abundanceplot(abund, xticks=(1:10, samplenames(abund)), xrotation=45)
+
+savefig("abundanceplot.png"); nothing # hide
+```
 
-```@repl 1
-using StatPlots
+![abundance plot](./abundanceplot.png)
 
-srand(1) # hide
+Perhaps you have some metadata that you'd like to add as well:
 
-abund = abundancetable(
-           rand(100, 10),
-           ["sample_$x" for x in 1:10],
-           ["feature_$x" for x in 1:100]);
-relativeabundance!(abund)
+```@example 1
+labels = ["a","a","b","a","b","b","b","b","a","a"]
 
-abundanceplot(abund)
+plot(
+    abundanceplot(abund, xticks=(1:10, samplenames(abund)), xrotation=45),
+    plot(annotationbar(labels)),
+    layout=grid(2,1, heights=[0.9,0.1]))
 
-savefig("abundanceplot.png"); nothing # hide
+savefig("abundanceplot-annotations.png"); nothing # hide
 ```
 
-![](abundanceplot.png)
+![abundance plot with annotations](./abundanceplot-annotations.png)

docs/src/distances.md

@@ -1,90 +1,76 @@
 # Working with Distances / Dissimilarity
 
-Quite often, it's useful to boil stuff down to distances between samples. For
-this, I'm using an interface with `Distances.jl` to generate a symetric
-`DistanceMatrix`, which also contains a vector for samples, and a field
-specifying which type of distance was used to calulate it. You can load one
-in manually, or generate it from an `AbundanceTable`.
+Quite often, it's useful to boil stuff down to distances between samples.
+`AbundanceTable`s can be used with the `pairwise()` function
+from [`Distances.jl`](https://github.com/JuliaStats/Distances.jl)
+to get a symetric distance matrix.
 
-```@repl 2
+```@example 2
 using Distances
 using Microbiome
 
 abund = abundancetable([1  3  0;
                         4  8  3;
                         5  0  4]);
 
-dm = getdm(abund, BrayCurtis())
+dm = pairwise(BrayCurtis(), abund, dims=2)
 ```
 
-I've also implemented a method to do a principle coordinates analysis. If
-necessary, you can include `correct_neg=true` to use the correction method
-described in [Lingoes (1971)](http://dx.doi.org/10.1007/BF02291398)
+To plot this, use the `MDS` or `PCA` implementations
+from [MultivariateStats](https://github.com/JuliaStats/MultivariateStats.jl) [^1]
+and plotting functionality
+from [StatsPlots](https://github.com/JuliaPlots/StatsPlots.jl)[^2].
 
-```@repl 2
-p = pcoa(dm)
+```@example 2
+using MultivariateStats
+using StatsPlots
 
-eigenvalue(p, 2)
-principalcoord(p, 1)
-variance(p, [1,2])
-```
-
-## Plotting
-
-**NOTE: The following functions are not currently working - I've moved them to a new package to simplify dependencies. I'm leaving the docs for now as a reference - see `Microbiome.jl` versions 0.2.1 and below for working versions**
-
-Some convenience plotting types are available using [`RecipesBase`](https://github.com/juliaplots/recipesbase.jl).
+mds = fit(MDS, dm, distances=true)
 
-```@repl 2
-using StatPlots
+plot(mds)
 
-srand(1) # hide
-abund = abundancetable(
-    rand(100, 10),
-    ["sample_$x" for x in 1:10],
-    ["feature_$x" for x in 1:100]);
-
-dm = getdm(abund, BrayCurtis());
-p = pcoa(dm, correct_neg=true);
-
-plot(p, title="Random PCoA")
-savefig("pcoplot.png"); nothing # hide
+savefig("mds.png"); nothing # hide
 ```
 
-![pcoa plot](pcoplot.png)
+![mds plot](./mds.png)
 
 ### Optimal Leaf Ordering
 
-I've also provided a plotting recipe for making treeplots for `Hclust` objects
-from the [`Clustering.jl`](http://github.com/JuliaStats/Clustering.jl) package:
+I also wrote a plotting recipe for making treeplots for `Hclust` objects
+from the [`Clustering.jl`](http://github.com/JuliaStats/Clustering.jl) package,
+and the recipe for plotting was moved into StatsPlots:
 
-```@repl 2
+```@example 2
 using Clustering
 
 dm = [0. .1 .2
       .1 0. .15
       .2 .15 0.];
 
-h = hclust(dm, :single);
-h.labels = ["a", "b", "c"];
+h = hclust(dm, linkage=:single);
 
-hclustplot(h)
+plot(h)
 savefig("hclustplot1.png"); nothing # hide
 ```
 
-![hclust plot 1](hclustplot1.png)
+![hclust plot 1](./hclustplot1.png)
 
 Note that even though this is a valid tree, the leaf `a` is closer to leaf `c`,
 despite the fact that `c` is more similar to `b` than to `a`. This can be fixed
 with a method derived from the paper:
 
-[Bar-Joseph et. al. "Fast optimal leaf ordering for hierarchical clustering." _Bioinformatics_. (2001)](https://doi.org/10.1093/bioinformatics/17.suppl_1.S22)
+[Bar-Joseph et. al. "Fast optimal leaf ordering for hierarchical clustering." _Bioinformatics_. (2001)](https://doi.org/10.1093/bioinformatics/17.suppl_1.S22)[^3]
 
-```@repl 2
-optimalorder!(h, dm)
-hclustplot(h)
+```@example 2
+h2 = hclust(dm, linkage=:single, branchorder=:optimal);
+
+plot(h2)
 
 savefig("hclustplot2.png"); nothing # hide
 ```
 
-![hclust plot 1](hclustplot2.png)
+![hclust plot 1](./hclustplot2.png)
+
+[^1]: Requires https://github.com/JuliaStats/MultivariateStats.jl/pull/85
+[^2]: Requires https://github.com/JuliaPlots/StatsPlots.jl/pull/152
+[^3]: Requires https://github.com/JuliaStats/Clustering.jl/pull/170

docs/src/index.md

@@ -1,4 +1,6 @@
-# Microbiome.jl <small>For analysis of microbiome and microbial community data</small>
+# Microbiome.jl
+
+## For analysis of microbiome and microbial community data
 
 [![Latest Release](https://img.shields.io/github/release/BioJulia/Microbiome.jl.svg)](https://github.com/BioJulia/Microbiome.jl/releases/latest)
 [![Microbiome](http://pkg.julialang.org/badges/Microbiome_0.6.svg)](http://pkg.julialang.org/?pkg=Microbiome)