1
1
mirror of https://github.com/kanaka/mal.git synced 2024-11-11 00:52:44 +03:00
Commit Graph

7 Commits

Author SHA1 Message Date
Joel Martin
1332215754 perf: more accurate iter count (last-iters).
We should return the iter count prior to then final iteration because
the final one was incomplete when the 10 seconds expired.
2018-07-16 12:33:36 -05:00
Joel Martin
5ffa3314f9 perf: simplify perf3 iter calc. Enable perf^mal.
Since the iter count increases by 1 each time through the "loop", when
we hit more than 10 seconds elapsed time we can just return the
previous iter count rather than doing a percentage calculation. This
fixes results for basic modes which couldn't do the percent
calculation because itermediate values were greater than 32,768 which
is the limit for basic MODES.

Also, the mal implementation now runs and returns legit results so
re-enable perf^impl.
2018-07-16 12:24:05 -05:00
Joel Martin
6da59ec2f4 perf: fix perf3 to show total iters instead of / 3
The run-fn-for function was originally name run-fn and was hard-coded
to run for 3 seconds. I forgot to replace the 3 with the max-secs
variable. However, given that some implementations have less than
1 iteration per second, I'm just going to drop that division and
report the iterations per max-seconds so that for the slowest
implementations we have some relative comparison instead of all of
them just reporting 0.

Thanks to Tim Morgan for discovering this bug:
699f0ad23a
2018-07-16 09:34:37 -05:00
Joel Martin
699f0ad23a Add long running perf test. 2015-02-28 10:38:17 -06:00
Joel Martin
7f567f3645 matlab: stepA, perf test, comments. 2015-02-09 00:23:10 -06:00
Joel Martin
f4c8a091aa Make: refactor number type. Support large numbers.
- Can support numbers up to 100 decimal digits in length.
- Still no support for negative numbers or floating point.
- Change time-secs to time-ms and remove conditional in perf.mal
2015-01-09 16:16:56 -06:00
Joel Martin
a77e2b31de rust: add step8_macros 2015-01-06 21:58:59 -06:00