marcpabst / anova.jl Goto Github PK

View Code? Open in Web Editor NEW

21.0 21.0 9.0 65 KB

Provides a Simple Way to Calculate ANOVAs From Fitted Linear Models.

License: Other

Julia 100.00%

anova.jl's People

Contributors

Stargazers

Watchers

Forkers

neukon epogrebnyak kleinschmidt jbrea babaq standardgalactic lindemann09

anova.jl's Issues

add ANOVA failed on julia v1.2

I tried to add ANOVA, but got this error:

ERROR: Unsatisfiable requirements detected for package ANOVA [0825541b]:
ANOVA [0825541b] log: ├─possible versions are: [0.0.1, 0.1.0] or uninstalled ├─restricted to versions * by an explicit requirement, leaving only versions [0.0.1, 0.1.0] ├─restricted by julia compatibility requirements to versions: 0.1.0 or uninstalled, leaving only versions: 0.1.0 └─restricted by compatibility requirements with StatsModels [3eaba693] to versions: 0.0.1 or uninstalled — no versions left
└─StatsModels [3eaba693] log: ├─possible versions are: [0.0.1-0.0.2, 0.1.0, 0.2.0-0.2.6, 0.3.0-0.3.1, 0.4.0, 0.5.0, 0.6.0-0.6.3] or uninstalled ├─restricted by compatibility requirements with DataFrames [a93c6f00] to versions: [0.1.0, 0.2.0-0.2.6, 0.3.0-0.3.1, 0.4.0, 0.5.0, 0.6.0-0.6.3] or uninstalled
│ └─DataFrames [a93c6f00] log: │ ├─possible versions are: [0.1.0, 0.2.0-0.2.5, 0.3.0-0.3.16, 0.4.0-0.4.3, 0.5.0-0.5.12, 0.6.0-0.6.11, 0.7.0-0.7.8, 0.8.0-0.8.5, 0.9.0-0.9.1, 0.10.0-0.10.1, 0.11.0-0.11.7, 0.12.0, 0.13.0-0.13.1, 0.14.0-0.14.1, 0.15.0-0.15.2, 0.16.0, 0.17.0-0.17.1, 0.18.0-0.18.4, 0.19.0-0.19.3] or uninstalled
│ ├─restricted to versions * by NeuroAnalysis [088af2c0], leaving only versions [0.1.0, 0.2.0-0.2.5, 0.3.0-0.3.16, 0.4.0-0.4.3, 0.5.0-0.5.12, 0.6.0-0.6.11, 0.7.0-0.7.8, 0.8.0-0.8.5, 0.9.0-0.9.1, 0.10.0-0.10.1, 0.11.0-0.11.7, 0.12.0, 0.13.0-0.13.1, 0.14.0-0.14.1, 0.15.0-0.15.2, 0.16.0, 0.17.0-0.17.1, 0.18.0-0.18.4, 0.19.0-0.19.3]
│ │ └─NeuroAnalysis [088af2c0] log: │ │ ├─possible versions are: 1.0.0 or uninstalled │ │ └─NeuroAnalysis [088af2c0] is fixed to version 1.0.0 │ └─restricted to versions 0.19.3 by an explicit requirement, leaving only versions 0.19.3 └─restricted by compatibility requirements with GLM [38e38edf] to versions: 0.6.0-0.6.3 └─GLM [38e38edf] log: ├─possible versions are: [0.2.0-0.2.5, 0.3.0-0.3.2, 0.4.0-0.4.8, 0.5.0-0.5.6, 0.6.0-0.6.1, 0.7.0, 0.8.0-0.8.1, 0.9.0, 0.10.0-0.10.1, 0.11.0, 1.0.0-1.0.2, 1.1.0-1.1.1, 1.2.0, 1.3.0-1.3.1] or uninstalled
├─restricted to versions * by NeuroAnalysis [088af2c0], leaving only versions [0.2.0-0.2.5, 0.3.0-0.3.2, 0.4.0-0.4.8, 0.5.0-0.5.6, 0.6.0-0.6.1, 0.7.0, 0.8.0-0.8.1, 0.9.0, 0.10.0-0.10.1, 0.11.0, 1.0.0-1.0.2, 1.1.0-1.1.1, 1.2.0, 1.3.0-1.3.1]
│ └─NeuroAnalysis [088af2c0] log: see above └─restricted to versions 1.3.1 by an explicit requirement, leaving only versions 1.3.1

is the v0.2 not updated on registry, or I have to clone it?

ANOVA for LinearMixedModels

@PharmCat has implemented ANOVA type III for LinearMixedModels in https://github.com/PharmCat/ReplicateBE.jl. Would it possible to have the functionality ported to this package?

Cannot Install - dependency issues

I tried to install this and got all sorts of dependency issues

julia> using ANOVA
 │ Package ANOVA not found, but a package named ANOVA is available from a registry. 
 │ Install package?
 │   (@v1.7) pkg> add ANOVA 
 └ (y/n) [y]: 
   Resolving package versions...
ERROR: Unsatisfiable requirements detected for package StatsBase [2913bbd2]:
 StatsBase [2913bbd2] log:
 ├─possible versions are: 0.24.0-0.33.16 or uninstalled
 ├─restricted to versions 0.30-0.33 by DiscriminantAnalysis [2091bfbe], leaving only versions 0.30.0-0.33.16
 │ └─DiscriminantAnalysis [2091bfbe] log:
 │   ├─possible versions are: 0.1.0 or uninstalled
 │   └─DiscriminantAnalysis [2091bfbe] is fixed to version 0.1.0
 ├─restricted by compatibility requirements with Lasso [b4fcebef] to versions: 0.24.0-0.32.2, leaving only versions: 0.30.0-0.32.2
 │ └─Lasso [b4fcebef] log:
 │   ├─possible versions are: 0.1.1-0.6.3 or uninstalled
 │   ├─restricted to versions * by an explicit requirement, leaving only versions 0.1.1-0.6.3
 │   └─restricted by compatibility requirements with Distributions [31c24e10] to versions: 0.1.1-0.5.2 or uninstalled, leaving only versions: 0.1.1-0.5.2
 │     └─Distributions [31c24e10] log:
 │       ├─possible versions are: 0.16.0-0.25.49 or uninstalled
 │       ├─restricted to versions * by an explicit requirement, leaving only versions 0.16.0-0.25.49
 │       ├─restricted by compatibility requirements with ANOVA [0825541b] to versions: 0.16.0-0.23.12
 │       │ └─ANOVA [0825541b] log:
 │       │   ├─possible versions are: 0.1.0 or uninstalled
 │       │   └─restricted to versions * by an explicit requirement, leaving only versions 0.1.0
 │       └─restricted by compatibility requirements with Lasso [b4fcebef] to versions: [0.16.0-0.21.12, 0.23.0-0.23.12]
 │         └─Lasso [b4fcebef] log: see above
 └─restricted by compatibility requirements with EvalMetrics [251d5f9e] to versions: 0.33.0-0.33.16 — no versions left
   └─EvalMetrics [251d5f9e] log:
     ├─possible versions are: 0.1.0-0.2.1 or uninstalled
     └─restricted to versions * by an explicit requirement, leaving only versions 0.1.0-0.2.1
Stacktrace:
  [1] propagate_constraints!(graph::Pkg.Resolve.Graph, sources::Set{Int64}; log_events::Bool)
    @ Pkg.Resolve /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Resolve/graphtype.jl:1063
  [2] propagate_constraints! (repeats 2 times)
    @ /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Resolve/graphtype.jl:1000 [inlined]
  [3] simplify_graph!(graph::Pkg.Resolve.Graph, sources::Set{Int64}; clean_graph::Bool)
    @ Pkg.Resolve /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Resolve/graphtype.jl:1519
  [4] simplify_graph! (repeats 2 times)
    @ /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Resolve/graphtype.jl:1519 [inlined]
  [5] resolve_versions!(env::Pkg.Types.EnvCache, registries::Vector{Pkg.Registry.RegistryInstance}, pkgs::Vector{Pkg.Types.PackageSpec}, julia_version::VersionNumber)
    @ Pkg.Operations /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Operations.jl:335
  [6] targeted_resolve(env::Pkg.Types.EnvCache, registries::Vector{Pkg.Registry.RegistryInstance}, pkgs::Vector{Pkg.Types.PackageSpec}, preserve::Pkg.Types.PreserveLevel, julia_version::VersionNumber)
    @ Pkg.Operations /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Operations.jl:1154
  [7] tiered_resolve(env::Pkg.Types.EnvCache, registries::Vector{Pkg.Registry.RegistryInstance}, pkgs::Vector{Pkg.Types.PackageSpec}, julia_version::VersionNumber)
    @ Pkg.Operations /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Operations.jl:1139
  [8] _resolve(io::Base.TTY, env::Pkg.Types.EnvCache, registries::Vector{Pkg.Registry.RegistryInstance}, pkgs::Vector{Pkg.Types.PackageSpec}, preserve::Pkg.Types.PreserveLevel, julia_version::VersionNumber)
    @ Pkg.Operations /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Operations.jl:1160
  [9] add(ctx::Pkg.Types.Context, pkgs::Vector{Pkg.Types.PackageSpec}, new_git::Set{Base.UUID}; preserve::Pkg.Types.PreserveLevel, platform::Base.BinaryPlatforms.Platform)
    @ Pkg.Operations /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/Operations.jl:1176
 [10] add(ctx::Pkg.Types.Context, pkgs::Vector{Pkg.Types.PackageSpec}; preserve::Pkg.Types.PreserveLevel, platform::Base.BinaryPlatforms.Platform, kwargs::Base.Pairs{Symbol, Base.TTY, Tuple{Symbol}, NamedTuple{(:io,), Tuple{Base.TTY}}})
    @ Pkg.API /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/API.jl:268
 [11] add(pkgs::Vector{Pkg.Types.PackageSpec}; io::Base.TTY, kwargs::Base.Pairs{Symbol, Union{}, Tuple{}, NamedTuple{(), Tuple{}}})
    @ Pkg.API /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/API.jl:149
 [12] add(pkgs::Vector{Pkg.Types.PackageSpec})
    @ Pkg.API /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/API.jl:144
 [13] #add#27
    @ /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/API.jl:142 [inlined]
 [14] add
    @ /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/API.jl:142 [inlined]
 [15] try_prompt_pkg_add(pkgs::Vector{Symbol})
    @ Pkg.REPLMode /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/Pkg/src/REPLMode/REPLMode.jl:710
 [16] #invokelatest#2
    @ ./essentials.jl:716 [inlined]
 [17] invokelatest
    @ ./essentials.jl:714 [inlined]
 [18] check_for_missing_packages_and_run_hooks(ast::Any)
    @ REPL /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/REPL/src/REPL.jl:175
 [19] eval_user_input(ast::Any, backend::REPL.REPLBackend)
    @ REPL /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/REPL/src/REPL.jl:145
 [20] repl_backend_loop(backend::REPL.REPLBackend)
    @ REPL /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/REPL/src/REPL.jl:244
 [21] start_repl_backend(backend::REPL.REPLBackend, consumer::Any)
    @ REPL /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/REPL/src/REPL.jl:229
 [22] run_repl(repl::REPL.AbstractREPL, consumer::Any; backend_on_current_task::Bool)
    @ REPL /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/REPL/src/REPL.jl:362
 [23] run_repl(repl::REPL.AbstractREPL, consumer::Any)
    @ REPL /Applications/Julia-1.7.app/Contents/Resources/julia/share/julia/stdlib/v1.7/REPL/src/REPL.jl:349
 [24] (::Base.var"#930#932"{Bool, Bool, Bool})(REPL::Module)
    @ Base ./client.jl:394
 [25] #invokelatest#2
    @ ./essentials.jl:716 [inlined]
 [26] invokelatest
    @ ./essentials.jl:714 [inlined]
 [27] run_main_repl(interactive::Bool, quiet::Bool, banner::Bool, history_file::Bool, color_set::Bool)
    @ Base ./client.jl:379
 [28] exec_options(opts::Base.JLOptions)
    @ Base ./client.jl:309
 [29] _start()
    @ Base ./client.jl:495

I know this isn't being actively developed, but it would be handy if it could be kept up-to-date
as far as loading.
Thanks

Interface without GLM

I still do not understand how to use GLM to do ANOVA. Nor do I understand what EffectsCoding() is, or what contrasts do in a call to fit.

So below is something I made - a simple interface to take a number of vectors as inputs, and preform anova on them, without making a GLM. My proposition is to include this interface in the package, for users like myself:

using PrettyTables, Statistics, Distributions
SS(inputs::AbstractArray) = sum((i - mean(inputs))^2 for i in inputs)
SSW(all_obs::AbstractVector...) = sum(SS.(all_obs))
SSB(all_obs::AbstractVector...) = sum(length(ys) * (mean(ys) - mean(vcat(all_obs...)))^2 for ys in all_obs)
function SST(all_obs::AbstractVector...)
    ys = vcat(all_obs...)
    sum((y - mean(ys))^2 for y in ys)
end

function anova1(all_obs_input::AbstractVector...)
    all_obs = all_obs_input .|> skipmissing .|> collect
    observations = vcat(all_obs...)
    SS = [SSB(all_obs...), SSW(all_obs...), SST(all_obs...)]
    df = [length(all_obs) - 1, length(observations) - length(all_obs), length(observations) - 1]
    S² = [SS[i] / df[i] for i = 1:3]
    TS = round(S²[1] / S²[2], digits = 5)
    dist = FDist(df[1], df[2])
    p_val = round(1 - cdf(dist, TS[1]), digits = 5)
    table = pretty_table(hcat([:Between, :Within, :Total], SS, df, S²),
        header = ["Variation", "SS", "df", "S²"],
        alignment = :c)
    println("Test statistic = $(TS)")
    println("p-value        = $(p_val)")
    return table
end

v1 = [missing, 4, 2, 4]
v2 = [-1, 0, missing, 1]
v3 = [-2, 1, 1, missing]

anova1(v1, v2, v3)
#returns
┌───────────┬─────────┬────┬─────────┐
│ Variation │   SS    │ df │   S²    │
├───────────┼─────────┼────┼─────────┤
│  Between  │ 22.2222 │ 2  │ 11.1111 │
│  Within   │ 10.6667 │ 6  │ 1.77778 │
│   Total   │ 32.8889 │ 8  │ 4.11111 │
└───────────┴─────────┴────┴─────────┘
Test statistic = 6.25   
p-value        = 0.03411

The following can be added to allow matrix inputs, similar to the API i encountered in MATLAB.

anova1(all_obs_input::AbstractMatrix) = anova1(Vector.(eachcol(all_obs_input))...)
anova1([v1 v2 v3])

If this is to be THE package for ANOVA in Julia, I think that an API without using GLM should be part of it, because
a) it is simpler to use, and does not require understanding of GLM
b) One can use GLM directly, should one already be fluent in its use, to do ANOVA. This package should have a low bar of entry - it is currently too high for me.

different 1-way anova result of type 3 vs type 1/2

I have a dataset with just 1 factor, following the example on ReadMe, I got this result with the default type 3 anova:

but with type 2, it show this, which is what I expected.

What are ANOVA types I-III?

Perhaps a bit more references in the readme could help new users, some suggestions are here. I'm not sure what are types I, II, III of ANOVA? What is the source of this terminology?

Docs: Examples of how to specify model

I know nothing about linear models, so the requirement to define a linear model is not clear. Could you please provide examples of how to specify models that will be recognized (factor, interaction term, etc.)? And in the case of a type 3 (which I see is cased in the code but the README says is forthcoming), how do you specify which factors are random or fixed?

Avoid using internals and rely on API

I encourage the implementation to be based solely on the StatsBase.StatisticalModel / StatsBase.RegressionModel API such that it works with any regression model rather than limiting it to only the GLM struct. As for the GLM, I believe it wasn't robust to allowrankdeficient which could benefit from the generalization and checks. If you need methods beyond the current API, do reach out and we can expand the API to accommodate it.

Compatibility with DataFrames >= 1

Would it be possible to add compatibility with the recently update DataFrames package?

When requesting to install DataFrames version 1 (which I need for other packages), I get:

julia> Pkg.add(Pkg.PackageSpec(;name="DataFrames", version="1.0"))
   Resolving package versions...
ERROR: Unsatisfiable requirements detected for package DataFrames [a93c6f00]:
 DataFrames [a93c6f00] log:
 ├─possible versions are: 0.11.7-1.2.2 or uninstalled
 ├─restricted to versions * by KmerTools [1bdae644], leaving only versions 0.11.7-1.2.2
 │ └─KmerTools [1bdae644] log:
 │   ├─possible versions are: 0.1.0 or uninstalled
 │   └─KmerTools [1bdae644] is fixed to version 0.1.0
 ├─restricted to versions 1.0 by an explicit requirement, leaving only versions 1.0.0-1.0.2
 └─restricted by compatibility requirements with ANOVA [0825541b] to versions: 0.11.7-0.21.8 — no versions left
   └─ANOVA [0825541b] log:
     ├─possible versions are: 0.1.0 or uninstalled
     └─restricted to versions * by an explicit requirement, leaving only versions 0.1.0

more convenient way to create ANOVAs (ezANOVA)

README.md states that "more convenient way to create ANOVAs (similar to the ez package in R) is planned." Are there still plans to do that? Is the package still maintained.

Type III with interactions across software

I am in need of the type III ANOVA à la SAS for GLM and MixedModels (with support for rank deficient). A similar tool is the car::Anova(::lmerMod, type = 3) function in R. I am not an expert in ANOVA since I tend to use the standard Wald test routine, but for some software I am developing it is standard to report that ANOVA test. I am willing to implement these given some guidance on the procedure. I can start with a PR to handle the rank deficient LinearModel. Given some guidance, I can also extend the functionality to the LinearMixedModel.

ArgumentError: F test is only valid for nested models

I try to do classic bioequivalence testing for crossover design and get this error. Can anova could be done for this?
used code:

olscmax  = lm(@formula(log(Cmax) ~ Sequence+Period+Formulation+Subject), df; dropcollinear = true)
Anova(olscmax)

Output of anova() prints twice at REPL

The output of anova() at the REPL prints its output normally first, then repeats it after the Julia> prompt interrupting the first line. Typing at the prompt moves the prompt to the last line, overwriting it.

This is in JuliaPro 1.0.1.1 with Julia 1.0.2

Incompatible with ModelFrame v"0.6.3"

julia> Anova(model)
ERROR: type ModelFrame has no field terms
Stacktrace:
 [1] getproperty(::Any, ::Symbol) at .\sysimg.jl:18
 [2] (::getfield(ANOVA, Symbol("##Anova#1#4")))(::Int64, ::Type, ::StatsModels.TableRegressionModel{LinearModel{GLM.LmResp{Array{Float64,1}},GLM.DensePredChol{Float64,LinearAlgebra.Cholesky{Float64,Array{Float64,2}}}},Array{Float64,2}}) at C:\Users\xxx\.julia\dev\ANOVA\src\ANOVA.jl:26
 [3] Anova(::StatsModels.TableRegressionModel{LinearModel{GLM.LmResp{Array{Float64,1}},GLM.DensePredChol{Float64,LinearAlgebra.Cholesky{Float64,Array{Float64,2}}}},Array{Float64,2}}) at C:\Users\xxx\.julia\dev\ANOVA\src\ANOVA.jl:20
 [4] top-level scope at none:0

julia> Pkg.installed()["StatsModels"]
v"0.6.3"