No description
- Python 100%
| README.md | ||
| requirements.txt | ||
| tactic-timing.py | ||
Coq tactic timing summary
Script that takes the output of Coq with some Ltac time commands and
summarizes by labels and succeed/fail status.
Demo
Stats
sum count mean max
tactic success
tc_solve True 142.033 29517 0.004812 0.908
wp_pures True 123.621 1688 0.073235 1.028
tc_solve False 25.462 37264 0.000683 0.199
iFrameHyp True 24.534 1744 0.014068 0.884
False 20.885 2235 0.009345 0.199
iFrame any True 16.433 268 0.061317 1.044
Notice that "iFrameHyp" has two consecutive lines, one for cases where it succeeded and the other for failures.
The label "iFrame any" has a space, which is handled correctly.
To get the above log, I wrapped the implementations of three tactics,
tc_solve, iFrameHyp, and iFrame with calls to time, then re-compiled.
This produces output with thousands of lines like:
Tactic call tc_solve ran for 0.017 secs (0.017u,0.s) (success)
Tactic call iFrameHyp ran for 0.018 secs (0.018u,0.s) (success)
Tactic call tc_solve ran for 0.01 secs (0.01u,0.s) (success)
Tactic call iFrameHyp ran for 0.01 secs (0.01u,0.s) (success)
Tactic call tc_solve ran for 0.005 secs (0.005u,0.s) (success)
Tactic call iFrameHyp ran for 0.005 secs (0.005u,0.s) (success)