website: <https://arrizza.com/python-accurate-timed-loop>
## Summary
This is a python module that provides a way to have an acccurate timed loop.
For example if you need to do an activity every 250ms +/-10ms, this loop will do that.
## Sample code
see sample.py for a full example
```python
import accurate_timed_loop
loop_delay = 0.250 # seconds
total_wait = 25.0 # seconds
for elapsed, start_time in accurate_timed_loop.accurate_wait(total_wait, loop_delay):
# ... do task every 250 mS
pass
```
## Scripts
* See [Quick Start](https://arrizza.com/user-guide-quick-start) for information on using scripts.
* See [xplat-utils submodule](https://arrizza.com/xplat-utils) for information on the submodule.
## Accuracy and Limitations
The sample.py does testing and shows that on Windows MSYS2 the std deviation error is roughly
4mS in a 250mS loop. This means that 95% of loops will be +/-8 mS of the requested loop_delay.
```text
expected elapsed diff1(ms) actual(s) diff2(ms)
1 0.000000 0.000000 0.000 0.000000 0.000
2 0.250000 0.257294 7.294 0.257294 7.294
<snip>
100 24.750000 24.764093 14.093 24.764093 14.093
101 25.000000 25.015579 15.579 25.015579 15.579
Stats:
loop count : 101 loops
Error Range : 0.000 to 24.406 mS
Error Stddev : 5.009 mS
Error Average : 8.863 mS
Recommended adj: 0.012200
sample rc=0
doit overall rc=0
```
This value is specific to Windows and to the PC that it is running on.
To make it more accurate for your PC and OS use the fixed_adjustment parameter.
Set it so the minimum and maximum are roughly symmetrical around 0.
The Stdev and Average error values at that point should be miminal.
```python
import accurate_timed_loop
loop_delay = 0.250 # seconds
total_wait = 25.0 # seconds
adj = 0.009228 # macOS
for elapsed, start_time in accurate_timed_loop.accurate_wait(total_wait, loop_delay, fixed_adjustment=adj):
# ... do task every 250 mS
pass
```
Notes:
* Re-run this several times, and tweak the fixed adjustment.
* The sample.py reports a "Recommended adj" that usually results in better accuracy.
* macOS and Ubuntu tend to be less variant than Windows
This report shows that std deviation is much better.
```text
expected elapsed diff1(ms) actual(s) diff2(ms)
1 0.000000 0.000000 0.000 0.000000 0.000
2 0.250000 0.251537 1.537 0.251537 1.537
<snip>
101 25.000000 24.989502 -10.498 24.989502 -10.498
102 25.250000 25.241386 -8.614 25.241386 -8.614
Stats:
loop count : 102 loops
Error Range : -9.228 to 5.864 mS
Error Stddev : 1.238 mS
Error Average : 4.953 mS
Recommended adj: 0.009228
sample rc=0
doit overall rc=0
```
Limitations:
* there is NO guarantee that the average error will always be that low or that consistent
* the following runs were on a macOS
```text
# === first run:
Stats:
loop count : 102 loops
Error Range : -9.486 to 4.613 mS
Error Stddev : 1.962 mS
Error Average : 5.775 mS
Recommended adj: 0.009486
# === second run:
Stats:
loop count : 102 loops
Error Range : -9.587 to 3.287 mS
Error Stddev : 2.163 mS
Error Average : 6.745 mS
Recommended adj: 0.009587
# === third run:
Stats:
loop count : 102 loops
Error Range : -9.472 to 6.782 mS
Error Stddev : 1.546 mS
Error Average : 5.597 mS
Recommended adj: 0.009472
# === fourth run:
Stats:
loop count : 101 loops
Error Range : -9.518 to 10.365 mS
Error Stddev : 1.865 mS
Error Average : 5.410 mS
Recommended adj: 0.009518
# === fifth run:
Stats:
loop count : 101 loops
Error Range : -9.369 to 13.726 mS
Error Stddev : 2.196 mS
Error Average : 5.614 mS
Recommended adj: 0.009369
```
* if you use the adj parameter the incoming "elapsed" parameter will not be after your expected delay.
For example these two came in:
* at 24.749 seconds instead of the expected 24.750 seconds
* at 24.999 seconds instead of the expected 25.000 seconds
```text
expected elapsed diff1(ms) actual(s) diff2(ms)
100 24.750000 24.749573 -0.427 24.749573 -0.427
101 25.000000 24.999601 -0.399 24.999601 -0.399
```
Raw data
{
"_id": null,
"home_page": "https://bitbucket.org/arrizza-public/accurate-timed-loop/src/master",
"name": "accurate-timed-loop",
"maintainer": null,
"docs_url": null,
"requires_python": null,
"maintainer_email": null,
"keywords": "accurate loop, utility",
"author": "JA",
"author_email": "cppgent0@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/b0/0d/81e3fed964890a56826f640369e03d8563126c82194831f1eed9e7070b3f/accurate_timed_loop-0.0.17.tar.gz",
"platform": null,
"description": "website: <https://arrizza.com/python-accurate-timed-loop>\n\n## Summary\n\nThis is a python module that provides a way to have an acccurate timed loop.\n\nFor example if you need to do an activity every 250ms +/-10ms, this loop will do that.\n\n## Sample code\n\nsee sample.py for a full example\n\n```python\nimport accurate_timed_loop\n\nloop_delay = 0.250 # seconds\ntotal_wait = 25.0 # seconds\nfor elapsed, start_time in accurate_timed_loop.accurate_wait(total_wait, loop_delay):\n # ... do task every 250 mS\n pass\n```\n\n## Scripts\n\n* See [Quick Start](https://arrizza.com/user-guide-quick-start) for information on using scripts.\n* See [xplat-utils submodule](https://arrizza.com/xplat-utils) for information on the submodule.\n\n## Accuracy and Limitations\n\nThe sample.py does testing and shows that on Windows MSYS2 the std deviation error is roughly\n4mS in a 250mS loop. This means that 95% of loops will be +/-8 mS of the requested loop_delay.\n\n```text\n expected elapsed diff1(ms) actual(s) diff2(ms)\n 1 0.000000 0.000000 0.000 0.000000 0.000\n 2 0.250000 0.257294 7.294 0.257294 7.294\n<snip>\n100 24.750000 24.764093 14.093 24.764093 14.093\n101 25.000000 25.015579 15.579 25.015579 15.579\n\n\nStats:\nloop count : 101 loops\nError Range : 0.000 to 24.406 mS\nError Stddev : 5.009 mS\nError Average : 8.863 mS\nRecommended adj: 0.012200\n sample rc=0\n doit overall rc=0\n```\n\nThis value is specific to Windows and to the PC that it is running on.\n\nTo make it more accurate for your PC and OS use the fixed_adjustment parameter.\nSet it so the minimum and maximum are roughly symmetrical around 0.\nThe Stdev and Average error values at that point should be miminal.\n\n```python\nimport accurate_timed_loop\n\nloop_delay = 0.250 # seconds\ntotal_wait = 25.0 # seconds\nadj = 0.009228 # macOS\nfor elapsed, start_time in accurate_timed_loop.accurate_wait(total_wait, loop_delay, fixed_adjustment=adj):\n # ... do task every 250 mS\n pass\n```\n\nNotes:\n\n* Re-run this several times, and tweak the fixed adjustment.\n* The sample.py reports a \"Recommended adj\" that usually results in better accuracy.\n* macOS and Ubuntu tend to be less variant than Windows\n\nThis report shows that std deviation is much better.\n\n```text\n expected elapsed diff1(ms) actual(s) diff2(ms)\n 1 0.000000 0.000000 0.000 0.000000 0.000\n 2 0.250000 0.251537 1.537 0.251537 1.537\n<snip>\n101 25.000000 24.989502 -10.498 24.989502 -10.498\n102 25.250000 25.241386 -8.614 25.241386 -8.614\n\n\nStats:\nloop count : 102 loops\nError Range : -9.228 to 5.864 mS\nError Stddev : 1.238 mS\nError Average : 4.953 mS\nRecommended adj: 0.009228\n sample rc=0\n doit overall rc=0\n```\n\nLimitations:\n\n* there is NO guarantee that the average error will always be that low or that consistent\n* the following runs were on a macOS\n\n```text\n# === first run:\nStats:\nloop count : 102 loops\nError Range : -9.486 to 4.613 mS\nError Stddev : 1.962 mS\nError Average : 5.775 mS\nRecommended adj: 0.009486\n\n# === second run:\nStats:\nloop count : 102 loops\nError Range : -9.587 to 3.287 mS\nError Stddev : 2.163 mS\nError Average : 6.745 mS\nRecommended adj: 0.009587\n\n# === third run:\nStats:\nloop count : 102 loops\nError Range : -9.472 to 6.782 mS\nError Stddev : 1.546 mS\nError Average : 5.597 mS\nRecommended adj: 0.009472\n\n# === fourth run:\nStats:\nloop count : 101 loops\nError Range : -9.518 to 10.365 mS\nError Stddev : 1.865 mS\nError Average : 5.410 mS\nRecommended adj: 0.009518\n\n# === fifth run:\nStats:\nloop count : 101 loops\nError Range : -9.369 to 13.726 mS\nError Stddev : 2.196 mS\nError Average : 5.614 mS\nRecommended adj: 0.009369\n```\n\n* if you use the adj parameter the incoming \"elapsed\" parameter will not be after your expected delay.\n For example these two came in:\n * at 24.749 seconds instead of the expected 24.750 seconds\n * at 24.999 seconds instead of the expected 25.000 seconds\n\n```text\n expected elapsed diff1(ms) actual(s) diff2(ms)\n100 24.750000 24.749573 -0.427 24.749573 -0.427\n101 25.000000 24.999601 -0.399 24.999601 -0.399\n```\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Accurate timed loop",
"version": "0.0.17",
"project_urls": {
"Download": "https://bitbucket.org/arrizza-public/accurate-timed-loop/get/master.zip",
"Homepage": "https://bitbucket.org/arrizza-public/accurate-timed-loop/src/master"
},
"split_keywords": [
"accurate loop",
" utility"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "b00d81e3fed964890a56826f640369e03d8563126c82194831f1eed9e7070b3f",
"md5": "3b16623597bd0a7fb87c6d1f702431bd",
"sha256": "f5e50e497325bc92a03c1bbc18acdc0a3a18d603d7c7a5ad36d982d6b145b63e"
},
"downloads": -1,
"filename": "accurate_timed_loop-0.0.17.tar.gz",
"has_sig": false,
"md5_digest": "3b16623597bd0a7fb87c6d1f702431bd",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 31356,
"upload_time": "2024-05-12T00:18:50",
"upload_time_iso_8601": "2024-05-12T00:18:50.030145Z",
"url": "https://files.pythonhosted.org/packages/b0/0d/81e3fed964890a56826f640369e03d8563126c82194831f1eed9e7070b3f/accurate_timed_loop-0.0.17.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-05-12 00:18:50",
"github": false,
"gitlab": false,
"bitbucket": true,
"codeberg": false,
"bitbucket_user": "arrizza-public",
"bitbucket_project": "accurate-timed-loop",
"lcname": "accurate-timed-loop"
}
JSON