Standardize naming and use master checkouts for workspace dependencies

.github/workflows/CI.yml

@@ -51,22 +51,15 @@ jobs:
           version: ${{ matrix.version }}
           arch: ${{ matrix.arch }}
       - uses: julia-actions/cache@v3
-      # ProximalSolvers, CovarianceSelection and the ProximalBase that
-      # CovarianceSelection needs (SliceSum, FiberSum) are not registered yet: all
-      # from their masters. Julia 1.11 and later take them from [sources] in
-      # Project.toml; Julia 1.10 ignores [sources], and resolves each Pkg.add with
-      # the whole project, so they are added in one call. Remove the step and
-      # [sources] once they are registered.
-      - name: Add the unregistered dependencies from GitHub on Julia 1.10
-        if: matrix.version == 'lts'
-        shell: julia --color=yes --project=. {0}
-        run: |
-          using Pkg
-          Pkg.add([PackageSpec(url="https://github.com/mlakolar/ProximalBase.jl", rev="master"),
-                   PackageSpec(url="https://github.com/mlakolar/ProximalSolvers.jl", rev="master"),
-                   PackageSpec(url="https://github.com/mlakolar/CovarianceSelection.jl", rev="master")])
+      - name: Check naming conventions
+        run: ${{ runner.os == 'Windows' && 'python' || 'python3' }} scripts/check_naming.py
+      - name: Resolve dependencies from master checkouts
+        run: julia --color=yes scripts/instantiate.jl
       - uses: julia-actions/julia-buildpkg@v1
-      - uses: julia-actions/julia-runtest@v1
+      - name: Test against dependency master checkouts
+        run: julia --color=yes scripts/test_master.jl
+        env:
+          JULIA_TEST_COVERAGE: "true"
       - uses: julia-actions/julia-processcoverage@v1
       - uses: codecov/codecov-action@v7
         with:

.gitignore

@@ -15,3 +15,9 @@
 /exper/sim*/*.pdf
 /benchmark/compare/*/Manifest*.toml
 /benchmark/compare/results/
+
+# Generated master dependency test environments
+.master-test-env/
+
+# Local Julia-minor environments are regenerated from master checkouts
+/Manifest-v*.toml

CONTRIBUTING.md

@@ -0,0 +1,21 @@
+# Development conventions
+
+Types and modules use UpperCamelCase. Descriptive functions, fields and keyword
+arguments use snake_case; mutating functions end in `!`. Retain standard numerical
+abbreviations, mathematical identifiers and names required by upstream interfaces.
+Run `python3 scripts/check_naming.py` to check new declarations.
+
+Workspace dependencies use sibling master checkouts, including transitive
+dependencies. Missing checkouts are cloned from master. Existing checkouts are
+never reset or silently updated. Package version numbering and registry changes
+are deferred.
+
+Run the complete tests from the repository root:
+
+```sh
+julia scripts/test_master.jl
+```
+
+The script prepares an isolated master-based environment with test extras on
+Julia 1.10 and newer, then runs the suite without changing package dependencies.
+For the simulation environment, use `julia scripts/instantiate.jl exper`.

Project.toml

@@ -15,9 +15,9 @@ SpecialFunctions = "276daf66-3868-5448-9aa4-cd146d93841b"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 
 [sources]
-CovarianceSelection = {url = "https://github.com/mlakolar/CovarianceSelection.jl", rev = "master"}
-ProximalBase = {url = "https://github.com/mlakolar/ProximalBase.jl", rev = "master"}
-ProximalSolvers = {url = "https://github.com/mlakolar/ProximalSolvers.jl", rev = "master"}
+CovarianceSelection = {path = "../CovarianceSelection.jl"}
+ProximalBase = {path = "../ProximalBase.jl"}
+ProximalSolvers = {path = "../ProximalOPT.jl"}
 
 [compat]
 Aqua = "0.8"

README.md

@@ -18,16 +18,16 @@ independent Gaussian samples.
     `ReducedOracleNormal`, and with bootstrap standard errors
     `SymmetricOracleBoot`, `AsymmetricOracleBoot`
 * Simultaneous confidence intervals for all entries by the bootstrap:
-  `bootstrap`, `simulCI`, `simulCIstudentized`
-* Support recovery: `supportEstimate` with `ANTSupport`, `BootStdSupport` or
+  `bootstrap`, `simul_ci`, `simul_ci_studentized`
+* Support recovery: `support_estimate` with `ANTSupport`, `BootStdSupport` or
   `DTraceValidationSupport`
 
 DiffPrecTest is one of five Julia packages for high-dimensional statistics,
 three for estimation and inference and two for the optimization beneath them:
 
 | Package | |
 |---|---|
-| [CoordinateDescent.jl](https://github.com/mlakolar/CoordinateDescent.jl) | the lasso and its paths; the group, square-root, scaled and heteroscedastic lasso; varying-coefficient models |
+| [HighDimensionalModels.jl](https://github.com/mlakolar/HighDimensionalModels.jl) | the lasso and its paths; the group, square-root, scaled and heteroscedastic lasso; varying-coefficient models |
 | [CovarianceSelection.jl](https://github.com/mlakolar/CovarianceSelection.jl) | Gaussian graphical models: the graphical lasso, CLIME and ACLIME, precision matrices by the scaled lasso, the fused graphical lasso, and the difference of two precision matrices |
 | **DiffPrecTest.jl** (this package) | confidence intervals and tests for the difference of two precision matrices |
 | [ProximalSolvers.jl](https://github.com/mlakolar/ProximalSolvers.jl) | the solvers: coordinate descent with penalty paths, screening and active sets; proximal gradient methods, proximal Newton and ADMM |
@@ -53,8 +53,8 @@ res.p .+ [-1.96, 1.96] .* res.std
 
 # simultaneous 95% confidence intervals for the entries of the lower triangle of Δ;
 # the off-diagonal entries are multiplied by √2
-straps, = bootstrap(X, Y; bootSamples=500)
-CI = simulCIstudentized(straps, 0.95)
+straps, = bootstrap(X, Y; boot_samples=500)
+CI = simul_ci_studentized(straps, 0.95)
 ```
 
 `Distributions` also exports `estimate`; when both packages are loaded, call
@@ -68,7 +68,40 @@ coordinate descent of [ProximalSolvers.jl](https://github.com/mlakolar/ProximalS
 on losses of [ProximalBase.jl](https://github.com/mlakolar/ProximalBase.jl):
 the direct difference loss of [CovarianceSelection.jl](https://github.com/mlakolar/CovarianceSelection.jl)
 and the inverse Hessian losses are its `KronQuadratic`. The functions of the
-two stages (`diffEstimation`, `invQSymHessian`, `invAsymHessianEstimation`,
-`reducedDiffEstimation`, `invHessianReduced`) take the solver options as a
+two stages (`diff_estimation`, `inv_q_sym_hessian`, `inv_asym_hessian_estimation`,
+`reduced_diff_estimation`, `inv_hessian_reduced`) take the solver options as a
 ProximalSolvers `CDSolver`; so are the node-wise lasso regressions of
 `SeparateNormal`.
+
+## Naming migration
+
+Types retain `UpperCamelCase`; descriptive functions, fields and keyword
+arguments use `snake_case`. Mathematical identifiers remain available.
+
+| Previous spelling | Canonical spelling |
+|---|---|
+| `diffEstimation` | `diff_estimation` |
+| `invAsymHessianEstimation`, `invQSymHessian`, `invHessianReduced` | `inv_asym_hessian_estimation`, `inv_q_sym_hessian`, `inv_hessian_reduced` |
+| `supportEstimate` | `support_estimate` |
+| `simulCI`, `simulCIstudentized` | `simul_ci`, `simul_ci_studentized` |
+| `computeSimulationResult` | `compute_simulation_result` |
+| `bootSamples`, `estimSupport` | `boot_samples`, `estim_support` |
+| `SimulationResult.lenCoverage` | `SimulationResult.len_coverage` |
+
+Previous public spellings are deprecated aliases for a transition period.
+Supplying both spellings of a keyword raises an error. Package versions are
+unchanged. Active environments use sibling master checkouts of
+CovarianceSelection, ProximalBase and ProximalSolvers through local sources.
+The pinned `benchmark/compare/old` environment remains a historical baseline.
+
+## Development tests
+
+From the repository root, run:
+
+```sh
+julia scripts/test_master.jl
+```
+
+The script prepares an isolated test environment using dependency master
+checkouts and test extras on Julia 1.10 and newer. See
+[CONTRIBUTING.md](CONTRIBUTING.md) for the naming policy and environment setup.

benchmark/compare/new/Project.toml

@@ -11,10 +11,10 @@ SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
 Statistics = "10745b16-79ce-11e8-11f9-7d13ad32a3b2"
 
 [sources]
-CovarianceSelection = {url = "https://github.com/mlakolar/CovarianceSelection.jl", rev = "master"}
+CovarianceSelection = {path = "../../../../CovarianceSelection.jl"}
 DiffPrecTest = {path = "../../.."}
-ProximalBase = {url = "https://github.com/mlakolar/ProximalBase.jl", rev = "master"}
-ProximalSolvers = {url = "https://github.com/mlakolar/ProximalSolvers.jl", rev = "master"}
+ProximalBase = {path = "../../../../ProximalBase.jl"}
+ProximalSolvers = {path = "../../../../ProximalOPT.jl"}
 
 [compat]
 CovarianceSelection = "0.1"

exper/Project.toml

@@ -17,10 +17,10 @@ StatsBase = "2913bbd2-ae8a-5f71-8c99-4fb6c76f3a91"
 
 # ProximalSolvers is listed for its [sources] entry, until it is registered
 [sources]
-CovarianceSelection = {url = "https://github.com/mlakolar/CovarianceSelection.jl", rev = "master"}
+CovarianceSelection = {path = "../../CovarianceSelection.jl"}
 DiffPrecTest = {path = ".."}
-ProximalBase = {url = "https://github.com/mlakolar/ProximalBase.jl", rev = "master"}
-ProximalSolvers = {url = "https://github.com/mlakolar/ProximalSolvers.jl", rev = "master"}
+ProximalBase = {path = "../../ProximalBase.jl"}
+ProximalSolvers = {path = "../../ProximalOPT.jl"}
 
 [compat]
 CSV = "1"

exper/README.md

@@ -20,17 +20,17 @@ the true support (`O-...`).
 
 ## Setup
 
-With Julia 1.11 or later:
+From the repository root, prepare the experiment environment on Julia 1.10
+or newer:
 
 ```sh
-julia --project=. -e 'using Pkg; Pkg.instantiate()'
+julia scripts/instantiate.jl exper
 ```
 
-`Project.toml` takes DiffPrecTest from this repository, and CovarianceSelection,
-ProximalSolvers and ProximalBase from their masters on GitHub, until they are
-registered. Julia 1.10 ignores these sources; there, run
-`using Pkg; Pkg.develop(path=".."); Pkg.add([PackageSpec(url="https://github.com/mlakolar/ProximalBase.jl", rev="master"), PackageSpec(url="https://github.com/mlakolar/ProximalSolvers.jl", rev="master"), PackageSpec(url="https://github.com/mlakolar/CovarianceSelection.jl", rev="master")])`
-once instead.
+The bootstrap loads DiffPrecTest from this repository and CovarianceSelection,
+ProximalSolvers and ProximalBase from sibling master checkouts, including their
+transitive workspace dependencies. Existing checkouts are never reset or
+silently updated.
 
 ## Running
 

exper/sim10/sim10.jl

@@ -53,7 +53,7 @@ Sx = Symmetric( X'X / n )
 Sy = Symmetric( Y'Y / n )
 
 
-@time boot_res, eS = bootstrap(X, Y)
+@time boot_res, e_s = bootstrap(X, Y)
 
 
-@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res eS
+@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res e_s

exper/sim10/sim10_o.jl

@@ -55,7 +55,7 @@ Sy = Symmetric( Y'Y / n )
 # oracle supports: Δ = 0, so every entry is refitted on its own
 eS_init = [Int64[] for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 
 @save "$(dir)/res_$(ip)_$(rep).jld2" boot_res

exper/sim10/sim10_o_power.jl

@@ -83,6 +83,6 @@ Sy = Symmetric( Y'Y / n )
 S = [tril_index(p, r, c) for c = 1:p for r = c:p if abs(Δ[r, c]) > 1e-4]
 eS_init = [copy(S) for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 @save "$(dir)/res_$(rep).jld2" boot_res

exper/sim11/sim11.jl

@@ -55,7 +55,7 @@ Sx = Symmetric( X'X / n )
 Sy = Symmetric( Y'Y / n )
 
 
-@time boot_res, eS = bootstrap(X, Y)
+@time boot_res, e_s = bootstrap(X, Y)
 
 
-@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res eS
+@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res e_s

exper/sim11/sim11_o.jl

@@ -57,7 +57,7 @@ Sy = Symmetric( Y'Y / n )
 # oracle supports: Δ = 0, so every entry is refitted on its own
 eS_init = [Int64[] for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 
 @save "$(dir)/res_$(ip)_$(rep).jld2" boot_res

exper/sim11/sim11_o_power.jl

@@ -89,6 +89,6 @@ Sy = Symmetric( Y'Y / n )
 S = [tril_index(p, r, c) for c = 1:p for r = c:p if abs(Δ[r, c]) > 1e-4]
 eS_init = [copy(S) for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 @save "$(dir)/res_$(rep).jld2" boot_res

exper/sim12/sim12_o_power.jl

@@ -72,6 +72,6 @@ Sy = Symmetric( Y'Y / n )
 S = [tril_index(p, r, c) for c = 1:p for r = c:p if abs(Δ[r, c]) > 1e-4]
 eS_init = [copy(S) for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 @save "$(dir)/res_$(rep).jld2" boot_res

exper/sim5/sim5a.jl

@@ -48,16 +48,16 @@ Sy = Symmetric( cov(Y) )
 Sxtest = Symmetric( cov(Xtest) )
 Sytest = Symmetric( cov(Ytest) )
 
-@time eS = DiffPrecTest.__initSupport(Sx, Sy, X, Y)
-@time eΔNormal, eΔNormalP, _ = supportEstimate(ANTSupport(), X, Y, τArr; estimSupport=eS)
-@time eΔBoot  , eΔBootP, _ = supportEstimate(BootStdSupport(), X, Y, τArr; estimSupport=eS)
+@time e_s = DiffPrecTest.__init_support(Sx, Sy, X, Y)
+@time eΔNormal, eΔNormalP, _ = support_estimate(ANTSupport(), X, Y, τArr; estim_support=e_s)
+@time eΔBoot  , eΔBootP, _ = support_estimate(BootStdSupport(), X, Y, τArr; estim_support=e_s)
 
 # compete method
 
 maxλ = maximum( abs.(Sx - Sy) )
 minλ = maxλ * 0.04
 Λarr = exp.(range(log(maxλ), log(minλ), length=50))
 
-eΔDTr, i2, iInf, loss2arr, lossInfarr = supportEstimate(DTraceValidationSupport(), Sx, Sy, Sxtest, Sytest, Λarr)
+eΔDTr, i2, i_inf, loss2arr, loss_infarr = support_estimate(DTraceValidationSupport(), Sx, Sy, Sxtest, Sytest, Λarr)
 
-@save "$(dir)/res_$(rep).jld2" eΔNormal eΔNormalP eΔBoot eΔBootP eS eΔDTr i2 iInf loss2arr lossInfarr
+@save "$(dir)/res_$(rep).jld2" eΔNormal eΔNormalP eΔBoot eΔBootP e_s eΔDTr i2 i_inf loss2arr loss_infarr

exper/sim5/sim5b.jl

@@ -48,16 +48,16 @@ Sy = Symmetric( cov(Y) )
 Sxtest = Symmetric( cov(Xtest) )
 Sytest = Symmetric( cov(Ytest) )
 
-@time eS = DiffPrecTest.__initSupport(Sx, Sy, X, Y)
-@time eΔNormal, eΔNormalP, _ = supportEstimate(ANTSupport(), X, Y, τArr; estimSupport=eS)
-@time eΔBoot  , eΔBootP, _ = supportEstimate(BootStdSupport(), X, Y, τArr; estimSupport=eS)
+@time e_s = DiffPrecTest.__init_support(Sx, Sy, X, Y)
+@time eΔNormal, eΔNormalP, _ = support_estimate(ANTSupport(), X, Y, τArr; estim_support=e_s)
+@time eΔBoot  , eΔBootP, _ = support_estimate(BootStdSupport(), X, Y, τArr; estim_support=e_s)
 
 # compete method
 
 maxλ = maximum( abs.(Sx - Sy) )
 minλ = maxλ * 0.04
 Λarr = exp.(range(log(maxλ), log(minλ), length=50))
 
-eΔDTr, i2, iInf, loss2arr, lossInfarr = supportEstimate(DTraceValidationSupport(), Sx, Sy, Sxtest, Sytest, Λarr)
+eΔDTr, i2, i_inf, loss2arr, loss_infarr = support_estimate(DTraceValidationSupport(), Sx, Sy, Sxtest, Sytest, Λarr)
 
-@save "$(dir)/res_$(rep).jld2" eΔNormal eΔNormalP eΔBoot eΔBootP eS eΔDTr i2 iInf loss2arr lossInfarr
+@save "$(dir)/res_$(rep).jld2" eΔNormal eΔNormalP eΔBoot eΔBootP e_s eΔDTr i2 i_inf loss2arr loss_infarr

exper/sim5/summarize.jl

@@ -38,8 +38,8 @@ TPR_tr = zeros(NUM_REP, numΛ)
 dtrace = DataFrame(selection=String[], rep=Int[], tpr=Float64[], fpr=Float64[], precision=Float64[])
 
 for (rep, f) in enumerate(files)
-    eΔNormal, eΔBoot, eΔDTr, i2, iInf = jldopen(joinpath(dir, f), "r") do file
-        read(file, "eΔNormal"), read(file, "eΔBoot"), read(file, "eΔDTr"), read(file, "i2"), read(file, "iInf")
+    eΔNormal, eΔBoot, eΔDTr, i2, i_inf = jldopen(joinpath(dir, f), "r") do file
+        read(file, "eΔNormal"), read(file, "eΔBoot"), read(file, "eΔDTr"), read(file, "i2"), read(file, "i_inf")
     end
 
     for j=1:numτ
@@ -58,7 +58,7 @@ for (rep, f) in enumerate(files)
         TPR_tr[rep, j] = tpr(cm)
     end
 
-    for (selection, i) in (("frobenius", i2), ("max", iInf))
+    for (selection, i) in (("frobenius", i2), ("max", i_inf))
         cm = getConfusionMatrix(Δ, eΔDTr[i])
         push!(dtrace, (selection, rep, tpr(cm), fpr(cm), precision(cm)))
     end

exper/sim8/collect_our.jl

@@ -18,11 +18,11 @@ NUM_REP = parse(Int, get(ENV, "NUM_REP", "1000"))
 covers(CI) = all(CI[:, 1] .<= 0 .<= CI[:, 2])
 
 function coverage(res::Vector{BootstrapEstimates}, α::Float64=0.95)
-    mean(covers(simulCI(r, α)) for r in res)
+    mean(covers(simul_ci(r, α)) for r in res)
 end
 
 function coverage_studentized(res::Vector{BootstrapEstimates}, α::Float64=0.95)
-    mean(covers(simulCIstudentized(r, α)) for r in res)
+    mean(covers(simul_ci_studentized(r, α)) for r in res)
 end
 
 # loads folderName_rep.jld2 for the available repetitions

exper/sim8/compute_coverage.jl

@@ -14,5 +14,5 @@ res = load(fName, "results")
 # the intervals cover Δ = 0 when every one of them contains 0
 covers(CI) = all(CI[:, 1] .<= 0 .<= CI[:, 2])
 
-@show mean(covers(simulCI(r, alpha)) for r in res)
-@show mean(covers(simulCIstudentized(r, alpha)) for r in res)
+@show mean(covers(simul_ci(r, alpha)) for r in res)
+@show mean(covers(simul_ci_studentized(r, alpha)) for r in res)

exper/sim8/sim8.jl

@@ -54,7 +54,7 @@ Sy = Symmetric( Y'Y / n )
 
 
 
-@time boot_res, eS = bootstrap(X, Y)
+@time boot_res, e_s = bootstrap(X, Y)
 
 
-@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res eS
+@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res e_s

exper/sim8/sim8_o.jl

@@ -54,6 +54,6 @@ Sy = Symmetric( Y'Y / n )
 # oracle supports: Δ = 0, so every entry is refitted on its own
 eS_init = [Int64[] for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 @save "$(dir)/res_$(ip)_$(rep).jld2" boot_res

exper/sim8/sim8_o_power.jl

@@ -83,6 +83,6 @@ Sy = Symmetric( Y'Y / n )
 S = [tril_index(p, r, c) for c = 1:p for r = c:p if abs(Δ[r, c]) > 1e-4]
 eS_init = [copy(S) for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 @save "$(dir)/res_$(rep).jld2" boot_res

exper/sim9/sim9.jl

@@ -51,7 +51,7 @@ Sx = Symmetric( X'X / n )
 Sy = Symmetric( Y'Y / n )
 
 
-@time boot_res, eS = bootstrap(X, Y)
+@time boot_res, e_s = bootstrap(X, Y)
 
 
-@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res eS
+@save "$(dir)/res_$(ip)_$(rep).jld2" boot_res e_s

exper/sim9/sim9_o.jl

@@ -54,7 +54,7 @@ Sy = Symmetric( Y'Y / n )
 # oracle supports: Δ = 0, so every entry is refitted on its own
 eS_init = [Int64[] for j=1:div((p + 1)*p, 2)]
 
-@time boot_res, eS = bootstrap(X, Y; estimSupport=eS_init)
+@time boot_res, e_s = bootstrap(X, Y; estim_support=eS_init)
 
 
 @save "$(dir)/res_$(ip)_$(rep).jld2" boot_res