Problem structures

[paraynaud]

(WIP) use a DataFrame to print it with SolverBenchmark (both .tex and .md).

The following script prints a markdown tabular.
It informs the partially-separable structure of problems used to produce profiles.

using PartiallySeparableSolvers, PartiallySeparableNLPModels
using ExpressionTreeForge, PartitionedStructures
using Printf, Statistics

n = 10
include("models.jl")

ps_problems

path = pwd() * "/dvpt/comparaison/results/problem_analysis/"

io = open(path*"table.md", "w+")
@printf(io, "%s |%s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s \n", "name", "n", "N", "M", "constant", "linear", "quadratic", "cubic", "general", "convex", "concave", "general", "mininimal element dimension", "mean element dimension", "maximal element dimension", "minimal elemental contribution (for 1 variables)", "mean elemental contribution (for 1 variables)", "maximal elemental contribution (for 1 variables)")
@printf(io, "----|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----|----|\n")

for nlp in ps_problems
  psnlp = PartiallySeparableNLPModel(nlp)
  part_data = psnlp.part_data

  n = get_n(part_data)
  name = nlp.meta.name * string(n)
  N = get_N(part_data)
  M = get_M(part_data)

  element_functions = part_data.vec_elt_fun

  element_function_types = (elt_fun -> elt_fun.type).(element_functions)
  constant = count(is_constant, element_function_types)
  linear = count(is_linear, element_function_types)
  quadratic = count(is_quadratic, element_function_types)
  cubic = count(is_cubic, element_function_types)
  general_type = count(is_more, element_function_types)

  element_function_convexity_status = (elt_fun -> elt_fun.convexity_status).(element_functions)
  convex = count(is_convex, element_function_convexity_status)
  concave = count(is_concave, element_function_convexity_status)
  general_convex = count(is_unknown, element_function_convexity_status)

  length_element_functions = (elt_fun -> length(elt_fun.variable_indices)).(element_functions)
  mean_length_element_functions = mean(length_element_functions)
  min_length_element_functions = minimum(length_element_functions)
  max_length_element_functions = maximum(length_element_functions)

  pv = part_data.pv
  component_list = PartitionedStructures.get_component_list(pv)
  length_by_variable = (elt_list_var -> length(elt_list_var)).(component_list)
  mean_length_variable = mean(length_by_variable)
  min_length_variable = minimum(length_by_variable)
  max_length_variable = maximum(length_by_variable)
  @printf(io, "%s |%s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s | %s \n",name, n, N, M, constant, linear, quadratic, cubic, general_type, convex, concave, general_convex, min_length_element_functions, mean_length_element_functions, max_length_element_functions, min_length_variable, mean_length_variable, max_length_variable)
end
close(io)

The result is

name	n	N	M	constant	linear	quadratic	cubic	general	convex	concave	general	mininimal element dimension	mean element dimension	maximal element dimension	minimal elemental contribution (for 1 variables)	mean elemental contribution (for 1 variables)	maximal elemental contribution (for 1 variables)
arglina10	10	20	2	0	0	20	0	0	20	0	0	10	10.0	10	20	20.0	20
arglinb10	10	20	20	0	0	20	0	0	20	0	0	10	10.0	10	20	20.0	20
arglinc10	10	19	19	1	0	18	0	0	19	1	0	0	7.578947368421052	8	0	14.4	18
arwhead10	10	27	3	9	9	0	0	9	27	18	0	0	1.0	2	2	2.7	9
bdqrtic10	10	12	2	0	0	6	0	6	12	0	0	1	3.0	5	1	3.6	6
beale2	2	3	3	0	0	0	0	3	0	0	3	2	2.0	2	3	3.0	3
biggs66	6	52	39	13	0	0	0	39	13	13	39	0	1.5	2	13	13.0	13
brownbs2	2	3	3	0	0	2	0	1	2	0	1	1	1.3333333333333333	2	2	2.0	2
brybnd10	10	10	6	0	0	0	0	10	0	0	10	2	5.4	7	2	5.4	7
chainwoo8	8	19	5	1	0	12	0	6	13	1	6	0	1.5789473684210527	2	2	3.75	6
chnrosnb10	10	10	2	0	0	9	0	1	9	0	1	1	1.9	10	1	1.9	2
cliff2	2	4	4	0	2	1	0	1	4	2	0	1	1.25	2	2	2.5	3
clplatea9	9	17	4	0	1	14	0	2	17	1	0	1	1.7058823529411764	2	0	3.2222222222222223	8
clplateb9	9	19	4	0	3	14	0	2	19	3	0	1	1.631578947368421	2	0	3.4444444444444446	8
clplatec9	9	18	6	0	2	8	0	8	18	2	0	1	1.6666666666666667	2	0	3.3333333333333335	8
cosine10	10	9	1	0	0	0	0	9	0	0	9	2	2.0	2	1	1.8	2
cragglvy10	10	20	5	0	0	4	0	16	12	0	8	1	1.6	2	2	3.2	4
curly10	10	10	10	0	0	0	0	10	0	0	10	1	5.5	10	1	5.5	10
curly1010	10	10	10	0	0	0	0	10	0	0	10	1	5.5	10	1	5.5	10
curly2010	10	10	10	0	0	0	0	10	0	0	10	1	5.5	10	1	5.5	10
curly3010	10	10	10	0	0	0	0	10	0	0	10	1	5.5	10	1	5.5	10
dixmaane9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaanf9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaang9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaanh9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaani9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaanj9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaank9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaanl9	9	27	15	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaanm9	9	27	20	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaann9	9	27	27	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaano9	9	27	27	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixmaanp9	9	27	27	1	0	12	0	14	10	1	17	0	1.5925925925925926	2	4	4.777777777777778	5
dixon3dq10	10	10	1	0	0	10	0	0	10	0	0	1	1.8	2	1	1.8	2
dqdrtic10	10	16	2	0	0	16	0	0	16	0	0	1	1.5	2	1	2.4	3
dqrtic10	10	10	10	0	0	0	0	10	10	0	0	1	1.0	1	1	1.0	1
edensch10	10	28	4	1	0	9	0	18	19	1	9	0	1.2857142857142858	2	2	3.6	4
eg210	10	10	2	0	0	0	0	10	0	0	10	1	1.8	2	1	1.8	9
engval110	10	27	3	9	9	0	0	9	27	18	0	0	1.0	2	1	2.7	3
errinros10	10	18	10	0	0	9	0	9	9	0	9	1	1.5	2	1	2.7	3
extrosnb10	10	2	2	0	0	1	0	1	1	0	1	1	5.5	10	1	1.1	2
fletcbv210	10	3	3	0	1	1	0	1	2	1	1	1	7.0	10	2	2.1	3
fletcbv310	10	2	2	0	0	1	0	1	1	0	1	10	10.0	10	2	2.0	2
freuroth10	10	18	2	0	0	0	0	18	0	0	18	2	2.0	2	2	3.6	4
gaussian3	3	15	15	0	0	0	0	15	0	0	15	3	3.0	3	15	15.0	15
genhumps10	10	18	2	0	0	9	0	9	9	0	9	2	2.0	2	2	3.6	4
genrose10	10	11	3	1	0	9	0	1	10	1	1	0	1.7272727272727273	10	1	1.9	2
genrose_nash10	10	11	3	1	0	9	0	1	10	1	1	0	1.7272727272727273	10	1	1.9	2
hs12	2	2	2	0	0	1	0	1	1	0	1	1	1.5	2	1	1.5	2
hs22	2	2	2	0	0	1	0	1	1	0	1	1	1.5	2	1	1.5	2
hs32	2	2	2	0	1	1	0	0	2	1	0	1	1.5	2	1	1.5	2
hs384	4	6	4	0	0	4	0	2	3	0	3	1	1.6666666666666667	2	2	2.5	3
hs42	2	2	2	0	1	0	1	0	1	1	1	1	1.0	1	1	1.0	1
hs455	5	2	2	1	0	0	0	1	1	1	1	0	2.5	5	1	1.0	1
hs52	2	5	5	1	2	1	0	1	4	3	1	0	1.2	2	3	3.0	3
indef10	10	2	2	0	0	0	0	2	0	0	2	10	10.0	10	2	2.0	2
jennrichsampson2	2	10	10	0	0	0	0	10	0	0	10	2	2.0	2	10	10.0	10
kowosb4	4	11	11	0	0	0	0	11	0	0	11	4	4.0	4	11	11.0	11
liarwhd10	10	20	2	0	0	10	0	10	10	0	10	1	1.45	2	2	2.9	11
meyer33	3	16	16	0	0	0	0	16	0	0	16	3	3.0	3	16	16.0	16
morebv10	10	10	10	0	0	0	0	10	0	0	10	2	2.8	3	2	2.8	3
ncb2031	31	46	5	22	1	0	1	22	44	23	2	0	1.9782608695652173	30	1	2.935483870967742	4
ncb20b20	20	42	4	20	1	0	0	21	41	21	1	0	1.4285714285714286	20	3	3.0	3
noncvxu210	10	20	2	0	0	10	0	10	10	0	10	2	2.8	3	4	5.6	6
noncvxun10	10	20	2	0	0	10	0	10	10	0	10	1	2.7	3	2	5.4	8
nondia10	10	2	2	0	0	1	0	1	1	0	1	1	5.5	10	1	1.1	2
nondquar10	10	10	2	0	0	2	0	8	10	0	0	2	2.8	3	2	2.8	9
osborne15	5	33	33	0	0	0	0	33	0	0	33	5	5.0	5	33	33.0	33
palmer5c6	6	12	12	0	0	12	0	0	12	0	0	6	6.0	6	12	12.0	12
penalty210	10	20	12	0	0	1	0	19	1	0	19	1	1.9	10	3	3.8	4
penalty310	10	10	6	1	0	5	0	4	6	1	4	0	4.4	10	3	4.4	5
powellsg8	8	8	3	0	0	6	0	2	8	0	0	2	2.0	2	2	2.0	2
quartc10	10	10	10	0	0	0	0	10	10	0	0	1	1.0	1	1	1.0	1
genrose10	10	11	3	1	0	9	0	1	10	1	1	0	1.7272727272727273	10	1	1.9	2
sbrybnd10	10	10	10	0	0	0	0	10	0	0	10	2	5.4	7	2	5.4	7
schmvett10	10	24	3	0	0	0	0	24	0	0	24	2	2.3333333333333335	3	2	5.6	7
scosine10	10	9	9	0	0	0	0	9	0	0	9	2	2.0	2	1	1.8	2
sinquad10	10	10	3	0	0	0	0	10	1	0	9	1	2.7	3	1	2.7	10
spmsrtls100	100	230	197	0	0	0	0	230	0	0	230	1	2.139130434782609	3	3	4.92	5
srosenbr10	10	10	2	0	0	5	0	5	5	0	5	1	1.5	2	1	1.5	2
tointgss10	10	8	1	0	0	0	0	8	0	0	8	3	3.0	3	1	2.4	3
tquartic10	10	9	2	0	0	1	0	8	1	0	8	1	1.8888888888888888	2	0	1.7	9
tridia10	10	10	10	0	0	10	0	0	10	0	0	1	1.9	2	1	1.9	2
vardim10	10	12	2	0	0	12	0	0	12	0	0	1	2.5	10	3	3.0	3
woods8	8	12	4	0	0	8	0	4	8	0	4	1	1.6666666666666667	2	2	2.5	3

It this what you expected @dpo?