Avatar for the predict-idlab user
predict-idlab
tsdownsample
BlogDocsChangelog

:floppy_disk: add Python downsamplers + test identical as Rust impl

#33Merged
Comparing
algo_testing_python
(
bd66a40
) with
main
(
a5dc007
)
CodSpeed Performance Gauge
-38%
Improvement
8
Regression
12
Untouched
370

Benchmarks

390 total
test_everynth[5,000-1,000,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-38%
77.8 µs125 µs
test_everynth[5,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-38%
78 µs124.9 µs
test_minmaxlttb_with_x[True-int32-100-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-28%
3.6 ms4.9 ms
test_everynth[1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-23%
56.3 µs72.7 µs
test_everynth[1,000-1,000,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-22%
56.7 µs72.9 µs
test_minmaxlttb_no_x[True-int64-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-16%
3.8 ms4.5 ms
test_minmaxlttb_no_x[True-float64-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-16%
3.8 ms4.5 ms
test_everynth[100-1,000,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-16%
49.8 µs59.2 µs
test_everynth[100-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-16%
49.7 µs59 µs
test_minmaxlttb_no_x[True-int32-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-13%
2.5 ms2.9 ms
test_minmaxlttb_no_x[False-float64-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-10%
3.9 ms4.3 ms
test_minmaxlttb_no_x[False-int64-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
-10%
3.9 ms4.3 ms
test_minmaxlttb_with_x[True-int64-100-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
×2.6
10.1 ms3.8 ms
test_m4_with_x[True-float64-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+26%
2 ms1.6 ms
test_m4_with_x[True-float32-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+26%
1.7 ms1.3 ms
test_minmaxlttb_with_x[True-int32-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+24%
28.4 ms22.9 ms
test_minmax_with_x[True-float64-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+23%
2.1 ms1.7 ms
test_minmax_with_x[True-int32-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+23%
2 ms1.6 ms
test_minmaxlttb_with_x[False-float32-1,000-1,000,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+13%
10.8 ms9.6 ms
test_minmaxlttb_with_x[False-int32-1,000-1,000,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+12%
10.9 ms9.7 ms
test_minmaxlttb_with_x[False-float64-1,000-1,000,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+10%
13.3 ms12.1 ms
test_minmaxlttb_with_x[False-int64-1,000-1,000,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+10%
13.3 ms12.1 ms
test_minmaxlttb_with_x[False-float32-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+8%
4 ms3.7 ms
test_minmaxlttb_with_x[False-float64-100-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+7%
1.6 ms1.5 ms
test_minmaxlttb_with_x[False-int32-1,000-100,000]
tests/benchmarks/test_downsamplers.py
CodSpeed Performance Gauge
+7%
4.2 ms3.9 ms

Commits

Click on a commit to change the comparison range
Base
main
a5dc007
-39.91%
Merge branch 'main' into algo_testing_python
a423a10
3 years ago
by jvdd
+4.9%
perf: optimize blocksize increment
bbe44ad
3 years ago
by jvdd
-0.71%
Merge branch 'main' into algo_testing_python
b9d42a1
3 years ago
by jvdd
+0.09%
:broom: remove unnecessary floor operation
ce1af5d
3 years ago
by jvdd
-42.96%
:see_no_evil: unsafe load
231666e
3 years ago
by jvdd
+42.98%
:see_no_evil: unsafe load 2
e30f62a
3 years ago
by jvdd
+0.07%
:broom: load unsafe instead of slice
09cf661
3 years ago
by jvdd
-0.06%
:pray: branchless parallel no x minmax
7f4b9db
3 years ago
by jvdd
-47.92%
:pray: update parallel impl
4c27811
3 years ago
by jvdd
+38.93%
:see_no_evil: revert to pre-allocation impl
db05099
3 years ago
by jvdd
+8.78%
:pray: more efficient offset calculation
7c54c8f
3 years ago
by jvdd
+0.31%
:pray: optimize block offset + unsafe load
4745be9
3 years ago
by jvdd
-27.79%
:see_no_evil: :pray:
f1b1ed1
3 years ago
by jvdd
-14.95%
:broom:
175acb4
3 years ago
by jvdd
+42.36%
:see_no_evil: revert to old code base
d58b357
3 years ago
by jvdd
-1.89%
:pen: review code
bd66a40
3 years ago
by jvdd
© 2026 CodSpeed Technology
Home Terms Privacy Docs