-
Notifications
You must be signed in to change notification settings - Fork 21
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #547 from JuliaRobotics/master
v0.8.4-rc1
- Loading branch information
Showing
38 changed files
with
1,472 additions
and
602 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,26 @@ | ||
| name: CompatHelper | ||
|
|
||
| on: | ||
| schedule: | ||
| - cron: '20 00 * * *' | ||
| issues: | ||
| types: [opened, reopened] | ||
|
|
||
| jobs: | ||
| CompatHelper: | ||
| runs-on: ${{ matrix.os }} | ||
| strategy: | ||
| matrix: | ||
| julia-version: [1.2.0] | ||
| julia-arch: [x86] | ||
| os: [ubuntu-latest] | ||
| steps: | ||
| - uses: julia-actions/setup-julia@latest | ||
| with: | ||
| version: ${{ matrix.julia-version }} | ||
| - name: Pkg.add("CompatHelper") | ||
| run: julia -e 'using Pkg; Pkg.add("CompatHelper")' | ||
| - name: CompatHelper.main() | ||
| env: | ||
| GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }} | ||
| run: julia -e 'using CompatHelper; CompatHelper.main()' |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -6,3 +6,5 @@ test/tmp/*.dot | |
| test/tmp/*.tex | ||
|
|
||
| Manifest.toml | ||
|
|
||
| dev | ||
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,98 @@ | ||
| # Showcasing the available analysis tools for the Bayes (Junction) tree. | ||
| # using Revise | ||
|
|
||
| using IncrementalInference | ||
| using DistributedFactorGraphs # For `isSolvable` function. | ||
| using Combinatorics # For creating the variable ordering `permutations`. | ||
| using SuiteSparse.CHOLMOD: SuiteSparse_long # For CCOLAMD constraints. | ||
| using Gadfly # For histogram and scatter plots. | ||
| Gadfly.set_default_plot_size(35cm,25cm) | ||
|
|
||
|
|
||
| latex_fonts = Theme(major_label_font="CMU Serif", major_label_font_size=16pt, | ||
| minor_label_font="CMU Serif", minor_label_font_size=14pt, | ||
| key_title_font="CMU Serif", key_title_font_size=12pt, | ||
| key_label_font="CMU Serif", key_label_font_size=10pt) | ||
| Gadfly.push_theme(latex_fonts) | ||
|
|
||
| # Get tree for each variable ordering in a factor graph. | ||
| fg = generateCanonicalFG_Kaess(graphinit=false) | ||
| all_trees = getAllTrees(deepcopy(fg)) | ||
|
|
||
| # scores stores: (tree key ID, nnz, cost fxn 1, cost fxn 2). | ||
| unsorted_scores = Vector{Tuple{Int, Float64, Float64, Float64}}() | ||
| for key in keys(all_trees) | ||
| e = all_trees[key] # (Bayes tree, var order, nnz | ||
| tree = e[1] # Get the Bayes tree. | ||
| cost1 = getTreeCost_01(tree) | ||
| cost2 = getTreeCost_02(tree) | ||
| push!(unsorted_scores, (key, e[3], cost1, cost2)) | ||
| end | ||
|
|
||
| # Sort them to make sure the keys are in order. | ||
| scores = sort(unsorted_scores) | ||
|
|
||
| # Separate scores into vectors for plotting. | ||
| all_nnzs = (x->(x[2])).(scores) | ||
| costs_01 = (x->(x[3])).(scores) | ||
| costs_02 = (x->(x[4])).(scores) | ||
|
|
||
| min_ids_02 = findall(x->x == minimum(costs_02), costs_02) | ||
| max_ids_02 = findall(x->x == maximum(costs_02), costs_02) | ||
|
|
||
| min_ids_nnz = findall(x->x == minimum(all_nnzs), all_nnzs) | ||
| max_ids_nnz = findall(x->x == maximum(all_nnzs), all_nnzs) | ||
|
|
||
| # Find the intersection between best on both rubrics (lower left quadrant). | ||
| best_ids = findall(x->x in min_ids_02, min_ids_nnz) | ||
| # Find good factorizations but bad trees (upper left quadrant). | ||
| bad_trees_good_mats_ids = findall(x->x in max_ids_02, min_ids_nnz) | ||
| # Find good trees with bad matrix factorizations (lower right quadrant). | ||
| good_trees_bad_mats_ids = min_ids_02[findall(x->x == maximum(all_nnzs[min_ids_02]), all_nnzs[min_ids_02])] | ||
|
|
||
| # Get AMDs variable ordering. | ||
| amd_ordering = getEliminationOrder(fg) | ||
| amd_tree = buildTreeFromOrdering!(deepcopy(fg), amd_ordering) | ||
| amd_tree_nnz = nnzTree(amd_tree) | ||
| amd_tree_cost02 = getTreeCost_02(amd_tree) | ||
|
|
||
| # Get CCOLAMD variable ordering. First bring in CCOLAMD. | ||
| include(normpath(Base.find_package("IncrementalInference"), "..", "ccolamd.jl")) | ||
| A, varsym, fctsym = getAdjacencyMatrixSparse(fg) | ||
| colamd_ordering = varsym[Ccolamd.ccolamd(A)] | ||
| colamd_tree = buildTreeFromOrdering!(deepcopy(fg), colamd_ordering) | ||
| colamd_tree_nnz = nnzTree(colamd_tree) | ||
| colamd_tree_cost02 = getTreeCost_02(colamd_tree) | ||
|
|
||
| # Now add the iSAM2 constraint. | ||
| cons = zeros(SuiteSparse_long, length(A.colptr) - 1) | ||
| cons[findall(x->x == :x3, varsym)[1]] = 1 # NOTE(tonioteran) hardcoded for Kaess' example. | ||
| ccolamd_ordering = varsym[Ccolamd.ccolamd(A, cons)] | ||
| ccolamd_tree = buildTreeFromOrdering!(deepcopy(fg), ccolamd_ordering) | ||
| ccolamd_tree_nnz = nnzTree(ccolamd_tree) | ||
| ccolamd_tree_cost02 = getTreeCost_02(ccolamd_tree) | ||
|
|
||
| # Plot data points and underlying histogram. | ||
| bincnt = 20 | ||
| layers = [] | ||
| push!(layers, Gadfly.layer(x=[amd_tree_nnz], | ||
| y=[amd_tree_cost02], | ||
| Theme(default_color=colorant"green"))) | ||
| push!(layers, Gadfly.layer(x=[colamd_tree_nnz], | ||
| y=[colamd_tree_cost02], | ||
| Theme(default_color=colorant"blue"))) | ||
| push!(layers, Gadfly.layer(x=[ccolamd_tree_nnz], | ||
| y=[ccolamd_tree_cost02], | ||
| Theme(default_color=colorant"red"))) | ||
| push!(layers, Gadfly.layer(x=all_nnzs, | ||
| y=costs_02, | ||
| Geom.hexbin(xbincount=bincnt, ybincount=bincnt))) | ||
| pl = Gadfly.plot(layers..., | ||
| Guide.xlabel("Number of non zeros [int]"), | ||
| Guide.ylabel("Tree cost [cfxn2]"), | ||
| Guide.manual_color_key("", | ||
| ["AMD", "COLAMD", "iSAM2"], | ||
| ["green", "blue", "red"])) | ||
|
|
||
| img = SVG("vo_cost_canon_kaess.svg", 6inch, 6inch) | ||
| Gadfly.draw(img, pl) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.