for v0.3.1: support for julia v1.6, DataFrames v1.0, etc. (#13)

thanks CompatHelper and github-actions!

.travis.yml

@@ -4,7 +4,7 @@ os:
   - linux
   - osx
 julia:
-  - 1.5
+  - 1.6
 
 notifications:
   email: false
@@ -25,7 +25,7 @@ coveralls: true
 jobs:
   include:
     - stage: "Documentation"
-      julia: 1.5
+      julia: 1.6
       os: linux
       script:
         - julia --project=docs/ -e 'using Pkg; Pkg.instantiate();

Project.toml

@@ -1,7 +1,7 @@
 name = "QuartetNetworkGoodnessFit"
 uuid = "1382f7fc-2744-4d9d-8ec6-1e3efdec0746"
 authors = ["Cecile Ane <[email protected]>"]
-version = "0.3.0"
+version = "0.3.1"
 
 [deps]
 CSV = "336ed68f-0bac-5ca0-87d4-7b16caf5d00b"
@@ -18,9 +18,9 @@ StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
 
 [compat]
 CSV = "0.4, 0.5, 0.6, 0.7, 0.8"
-DataFrames = "0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.20, 0.21, 0.22"
+DataFrames = "0.13, 0.14, 0.15, 0.16, 0.17, 0.18, 0.19, 0.20, 0.21, 0.22, 1.0"
 NLopt = "0.5.1, 0.6"
-PhyloNetworks = "0.11, 0.12, 0.13"
+PhyloNetworks = "0.11, 0.12, 0.13, 0.14"
 SpecialFunctions = "0.8, 0.9, 0.10, 1.0"
 StatsFuns = "0.7, 0.8, 0.9"
-julia = "1.2, 1.3, 1.4, 1.5"
+julia = "1.2, 1.3, 1.4, 1.5, 1.6"

docs/Project.toml

@@ -6,6 +6,5 @@ DocumenterMarkdown = "997ab1e6-3595-5248-9280-8efb232c3433"
 PhyloNetworks = "33ad39ac-ed31-50eb-9b15-43d0656eaa72"
 
 [compat]
-CSV = "0.7"
-DataFrames = "0.21"
+CSV = "0.8"
 Documenter = "~0.26"

docs/src/man/gof.md

@@ -13,7 +13,7 @@ proportion of genes estimated to have each 4-taxon unrooted topology.
 
 ```@repl gof
 using QuartetNetworkGoodnessFit, DataFrames, CSV
-qCF = DataFrame!(CSV.File(joinpath(dirname(pathof(QuartetNetworkGoodnessFit)), "..","test","example_qCF_5taxa.csv")));
+qCF = DataFrame(CSV.File(joinpath(dirname(pathof(QuartetNetworkGoodnessFit)), "..","test","example_qCF_5taxa.csv")), copycols=false);
 qCF
 ```
 
@@ -45,7 +45,7 @@ lengths in the network, in coalescent units, before quantifying the
 goodness-of-fit.
 
 ```@repl gof
-res0 = quarnetGoFtest!(net0, qCF, true; seed=234, nsim=2);
+res0 = quarnetGoFtest!(net0, qCF, true; seed=201, nsim=3);
 nothing # hide
 ```
 
@@ -58,7 +58,7 @@ net0 = res0[5]
 Now we re-run the test using the option `false` to not re-optimize
 branch lengths. We use `nsim=200` simulations below to make
 this example faster. For a real data analysis, delete the `nsim` option
-to use the default instead (1000) or specify higher value.
+to use the default instead (1000) or specify a higher value.
 
 ```@repl gof
 res0 = quarnetGoFtest!(net0, qCF, false; seed=234, nsim=200);

test/Project.toml

@@ -4,7 +4,3 @@ DataFrames = "a93c6f00-e57d-5684-b7b6-d8193f3e46c0"
 Distributed = "8ba89e20-285c-5b6f-9357-94700520ee1b"
 PhyloNetworks = "33ad39ac-ed31-50eb-9b15-43d0656eaa72"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
-
-[compat]
-CSV = "0.7"
-DataFrames = "0.21"

test/test_qnetGoF.jl

@@ -1,6 +1,6 @@
 @testset "testing GoF, multinomial distribution" begin
 
-df = DataFrame!(CSV.File(joinpath(dirname(Base.find_package("PhyloNetworks")),"..","examples","buckyCF.csv")))
+df = DataFrame(CSV.File(joinpath(dirname(Base.find_package("PhyloNetworks")),"..","examples","buckyCF.csv")), copycols=false)
 d0 = readTableCF(df)
 d = deepcopy(d0)
 net3 = readTopology("((((D:0.4,C:0.4):4.8,((A:0.8,B:0.8):2.2)#H1:2.2::0.7):4.0,(#H1:0::0.3,E:3.0):6.2):2.0,O:11.2);");
@@ -63,12 +63,13 @@ Distributed.addprocs(2)
 # start with: julia -p 2 --project
 # or: using Distributed; @everywhere begin; using Pkg; Pkg.activate("."); using PhyloNetworks; end
 @everywhere using QuartetNetworkGoodnessFit
-netresult1 = quarnetGoFtest!(net3,d,false; seed=1456, nsim=5);
+netresult1 = quarnetGoFtest!(net3,d,false; seed=2298, nsim=5);
 @test netresult1[4] ≈ [0.0024449826689709165,0.01496306673600063,0.01496306673600063,0.0024449826689709165,0.04086460431063039,0.9998541057240138,0.1901450501005025,0.8909735618259936,0.9058717147295428,0.8909735618259936,0.1901450501005025,0.9058717147295428,0.9913859984840471,0.3656465603640152,0.04086460431063039]
 @test netresult1[2] ≈ 6.21966321647047 # z stat, uncorrected
 @test netresult1[3] ≈ 3.405362128771355 # sigma
 @test netresult1[6] ≈ vcat(7.4043609719886545, repeat([-0.8885233166386386],4))
-netresult1 = quarnetGoFtest!(net3,d,true; seed=182, nsim=2, quartetstat=:Qlog);
+netresult1 = (@test_logs (:warn, r"far from 0") quarnetGoFtest!(net3,d,true; seed=182, nsim=2, quartetstat=:Qlog));
+# just because 2 simulated z's only, and same values bc tiny network. may break with different RNG
 # note: with verbose=true, we see hybrid-lambda's warnings:
 # WARNING! NOT ULTRAMETRIC!!!
 # WARNING: Gene tree is not ultrametric

test/test_ticr.jl

@@ -1,7 +1,7 @@
 @testset "testing TICR, Dirichlet distribution" begin
 # previously in PhyloNetworks
 
-df = DataFrame!(CSV.File(joinpath(dirname(Base.find_package("PhyloNetworks")),"..","examples","buckyCF.csv")));
+df = DataFrame(CSV.File(joinpath(dirname(Base.find_package("PhyloNetworks")),"..","examples","buckyCF.csv")), copycols=false);
 d = readTableCF(df);
 
 @testset "ticr! on data frame, tree" begin
@@ -23,7 +23,7 @@ result3 = ticr!(net2_1,d,false);
 @test result3[1] ≈ 0.06932031690660927 # p-value, from R
 @test result3[3] == Dict("[0.0, 0.01)"=>2, "[0.01, 0.05)"=>0, "[0.05, 0.1)"=>2, "[0.1, 1.0]"=>11)
 @test result3[4][2] ≈ 54.241562916699216 # pseudo log-lik obtained from R
-@test result3[4][1] ≈ 20.128258663235194 # alpha obtained from R
+@test result3[4][1] ≈ 20.128258663235194 atol=1e-5 # alpha obtained from R
 result3_1 = ticr!(net2_1,d,false; test=:goodness);
 @test result3_1[2] ≈ 25.962962962962965463 # chi-squared statistic obtained from R
 @test result3_1[1] ≈ 9.7092282251534852702e-06 # p-value obtained from R