Skip to content

recovered regr_smlp/data/smlp_toy_num_resp_noknobs.csv and regr_smlp/specs/smlp_toy_num_resp_noknobs_verify.spec#151

Open
zurabksmlp wants to merge 20 commits into
masterfrom
noknobs_tests_recovery
Open

recovered regr_smlp/data/smlp_toy_num_resp_noknobs.csv and regr_smlp/specs/smlp_toy_num_resp_noknobs_verify.spec#151
zurabksmlp wants to merge 20 commits into
masterfrom
noknobs_tests_recovery

Conversation

@zurabksmlp
Copy link
Copy Markdown
Collaborator

Recovered missing data and spec files for the "noknobs" tests:
57,61,62,65,66,67,68,71,72,73,74,75,76,78,84,139,144,162,163,184,185,186
All tests use the same data and spec files:
data:
regr_smlp/data/smlp_toy_num_resp_noknobs
specL
regr_smlp/specs/smlp_toy_num_resp_noknobs_verify.spec

These data and spec files are obtained simply by renaming variables: : x --> x0, p1 --> x1, p2 --> x3 irrespectively n:
regr_smlp/data/smlp_toy_num_resp_mult.csv
regr_smlp/specs/smlp_toy_num_resp_mult_verify.spec

I have checked these changes on Test57 and have a good confidence that the recovered files are as intended.
This addresses part of missing input files issue #147

…specs/smlp_toy_num_resp_noknobs_verify.spec
@zurabksmlp zurabksmlp requested a review from mdmitry1 May 31, 2026 10:46
Copy link
Copy Markdown
Collaborator

@mdmitry1 mdmitry1 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

  1. There are many tests failed due to missing data:
echo n | env CUDA_VISIBLE_DEVICES=-1 python3.11 ./smlp_regr.py -t all -w 8 -def n -g |& tee noknobs_tests_recovery.smlp_regr.log
grep Failed noknobs_tests_recovery.smlp_regr.log
Test 73 Failed:
Test 74 Failed:
Test 75 Failed:
Test 139 Failed:
Test 162 Failed:
Test 163 Failed:
Test 184 Failed:
Test 185 Failed:
Test 186 Failed:
  1. Master is not updated

2.1 Missing files in master

grep "File master" noknobs_tests_recovery.smlp_regr.log | awk '{print $3}' | cut -d_ -f1 | sort -u
Test144
Test57
Test61
Test62
test65
Test65
Test66
test67
Test67
Test68
test71
Test71
Test72
test76
Test76
test78
Test78
Test84

2.2 Diff(s) due smlp package structure change.

For example:

cat test76_model_smlp_model_term.json | tr ' ' '\012' | sort > test76_model_smlp_model_term_sorted.json
cat ../master/test76_model_smlp_model_term.json | tr ' ' '\012' | sort > test76_model_smlp_model_term_sorted_master.json
diff test76_model_smlp_model_term_sorted*.json
215,216c215,216
< <smlp.core.libsmlp.term2
< <smlp.core.libsmlp.term2
---
> <smlp.libsmlp.term2
> <smlp.libsmlp.term2

My results: code.tar.gz

Version validated:

pip show smlptech | grep Version
Version: 1.2.0

@fbrausse
Copy link
Copy Markdown
Collaborator

fbrausse commented Jun 1, 2026

Checking this on my pytest branch we go from

= 31 failed, 163 passed, 38 skipped, 2 xfailed, 41 warnings in 558.42s (0:09:18) =

to

= 30 failed, 174 passed, 28 skipped, 2 xfailed, 32 warnings in 524.69s (0:08:44) =

Warnings removed concern path ../data/smlp_toy_num_resp_noknobs.csv (as expected for this PR), while new warnings are about:

  • -new_dat ../data/smlp_toy_num_resp_noknobs_pred_labeled.csv not existing (tests 73, 139, 162, 163, 184, 185, 186), and
  • -save_model t specified in args but no -model_name given (test 84).

@mdmitry1
Copy link
Copy Markdown
Collaborator

mdmitry1 commented Jun 1, 2026

Results after 0459981 checkin.
Regression command:

cd $(git rev-parse --show-toplevel)/tests/smlp_regression
./run_smlp_regression
  1. All inputs are in place
grep -c Failed run_smlp_regression.log
0
  1. Missing outputs run_smlp_regression.log
grep "File master" run_smlp_regression.log | awk '{print $3}' | cut -d_ -f1 | sort -u
Test139
Test144
Test162
Test163
Test184
Test185
Test186
Test61
Test62
Test65
test67
Test67
Test68
test71
Test71
Test72
test73
Test73
Test74
Test75
test76
Test76
test78
Test78
Test84
  1. Diffs: run_smlp_regression_diff_report.log

  2. All results: noknobs_tests_recovery_code.tar.gz

@fbrausse
Copy link
Copy Markdown
Collaborator

fbrausse commented Jun 1, 2026

Now even better:

= 33 failed, 180 passed, 19 skipped, 2 xfailed, 23 warnings in 546.12s (0:09:06) =

The 3 tests now failing that are new in wrt. before are 57, 65 and 66, due to

Details

Test57:

--- ../master/Test57_smlp_toy_num_resp_noknobs_verify_results.json	2026-06-01 14:02:21.048730624 +0200
+++ /tmp/pytest-of-kane/pytest-9/popen-gw13/test3/Test57_smlp_toy_num_resp_noknobs_verify_results.json	2026-06-01 14:03:44.418788290 +0200
@@ -4,10 +4,10 @@
 		"assertion_status": "FAIL",
 		"counter_example": {
 			"x0": 0.0,
-			"x1": 1.0,
+			"x1": 4.0,
 			"y1": 9.0,
-			"x2": 7.0,
-			"y2": 5.0
+			"x2": 6.000000067055225,
+			"y2": 9.0
 		},
 		"assertion_feasible": false
 	},

Test65

--- ../master/Test65_smlp_toy_num_resp_noknobs_verify_results.json	2026-06-01 14:02:21.054730700 +0200
+++ /tmp/pytest-of-kane/pytest-9/popen-gw6/test3/Test65_smlp_toy_num_resp_noknobs_verify_results.json	2026-06-01 14:03:49.898857820 +0200
@@ -10,7 +10,7 @@
 		"assertion_status": "FAIL",
 		"counter_example": {
 			"x0": 0.0,
-			"x1": 7.0,
+			"x1": 4.0,
 			"y1": 9.0,
 			"x2": 6.000000067055225,
 			"y2": 9.0

Test66

--- ../master/Test66_test65_model_trace.csv	2026-06-01 14:02:21.054730700 +0200
+++ /tmp/pytest-of-kane/pytest-9/popen-gw6/test4/Test66_test65_model_trace.csv	2026-06-01 14:04:04.129038377 +0200
@@ -1,9 +1,9 @@
 stage,solver,x0,x1,x2,y1,y2
 interface_consistency,sat,0,7,3
-model_consistency,sat,0,1,805306377/134217728,5,5
-witness_consistency,sat,0,1,805306377/134217728,5,5
-witness_consistency,sat,0,1,805306377/134217728,5,5
-ca,sat,0,1,6,5,5
+model_consistency,sat,0,7,805306377/134217728,9,9
+witness_consistency,sat,0,7,805306377/134217728,9,9
+witness_consistency,sat,0,7,805306377/134217728,9,9
+ca,sat,1,1,27/4,9,5
 ce,unsat
-ca,sat,0,7,671088649/134217728,5,9
-ce,sat,0,7,805306377/134217728,9,9
+ca,sat,0,1,1140850697/201326592,5,5
+ce,sat,0,4,805306377/134217728,9,9
--- ../master/Test66_test65_model_verify_results.json	2026-06-01 14:02:21.054730700 +0200
+++ /tmp/pytest-of-kane/pytest-9/popen-gw6/test4/Test66_test65_model_verify_results.json	2026-06-01 14:04:04.133038427 +0200
@@ -10,7 +10,7 @@
 		"assertion_status": "FAIL",
 		"counter_example": {
 			"x0": 0.0,
-			"x1": 7.0,
+			"x1": 4.0,
 			"y1": 9.0,
 			"x2": 6.000000067055225,
 			"y2": 9.0

@mdmitry1
Copy link
Copy Markdown
Collaborator

mdmitry1 commented Jun 1, 2026

Reproduced @fbrausse results for Test57:

diff Test57_smlp_toy_num_resp_noknobs_verify_results.json ../master/
7c7
< 			"x1": 4.0,
---
> 			"x1": 1.0,
9,10c9,10
< 			"x2": 6.000000067055225,
< 			"y2": 9.0
---
> 			"x2": 7.0,
> 			"y2": 5.0
image

@mdmitry1
Copy link
Copy Markdown
Collaborator

mdmitry1 commented Jun 2, 2026

Results after commit 88cdb5f Mon Jun 1 23:19:50 2026 +0300

code.88cdb5f.tar.gz

All inputs are in place

grep -c Failed noknobs_tests_recovery.diff_report.log
0

Missing master files

grep "File master" noknobs_tests_recovery.smlp_regr.log | awk '{print $3}' | cut -d_ -f1 | sort -u
Test61
Test62
Test71

Diffs

$(git rev-parse --show-toplevel)/tests/smlp_regression/create_diff_report | grep Diff | awk '{print $5}' | awk -F_ '{print $1}' | tr -d Test | sort -un | nl
     1	57
     2	65
     3	66
     4	67
     5	68
     6	72
     7	73
     8	74
     9	75
    10	76
    11	78
    12	139
    13	144
    14	162
    15	163
    16	184
    17	185
    18	186

Results match exactly on two machines

  • Ubuntu 24.04
  • Windows 11/WSL2 with Ubuntu 24.04

SMLP packages

smlp_packages.txt

smlp_dependencies_tree.txt

@fbrausse
Copy link
Copy Markdown
Collaborator

fbrausse commented Jun 2, 2026

Filtering out one more expected-to-fail test (229) as well as the three tests 145, 146 and 205 (for the reason mentioned in #149), pytest now arrives at

= 43 failed, 166 passed, 22 skipped, 3 xfailed, 26 warnings in 543.53s (0:09:03) =

Copy link
Copy Markdown
Collaborator

@mdmitry1 mdmitry1 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fixed all diffs

@mdmitry1 mdmitry1 requested a review from fbrausse June 2, 2026 17:17
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants