-
-
Notifications
You must be signed in to change notification settings - Fork 48
Fixed merging float definitions with different precisions #1577
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Flow PHP - BenchmarksResults of the benchmarks from this PR are compared with the results from 1.x branch. Extractors+-----------------------+-------------------+------+-----+-----------------+------------------+-----------------+
| benchmark | subject | revs | its | mem_peak | mode | rstdev |
+-----------------------+-------------------+------+-----+-----------------+------------------+-----------------+
| CSVExtractorBench | bench_extract_10k | 1 | 3 | 4.883mb +1.48% | 611.149ms -5.71% | ±0.40% -3.12% |
| JsonExtractorBench | bench_extract_10k | 1 | 3 | 4.957mb +1.46% | 1.140s +0.48% | ±0.53% -47.49% |
| ParquetExtractorBench | bench_extract_10k | 1 | 3 | 86.470mb +0.02% | 917.708ms -0.28% | ±0.27% -71.72% |
| TextExtractorBench | bench_extract_10k | 1 | 3 | 4.606mb +0.15% | 38.654ms -1.06% | ±0.91% +9.37% |
| XmlExtractorBench | bench_extract_10k | 1 | 3 | 4.581mb +0.13% | 605.292ms +0.15% | ±1.07% +167.52% |
+-----------------------+-------------------+------+-----+-----------------+------------------+-----------------+
Transformers+-----------------------------+--------------------------+------+-----+------------------+-----------------+----------------+
| benchmark | subject | revs | its | mem_peak | mode | rstdev |
+-----------------------------+--------------------------+------+-----+------------------+-----------------+----------------+
| RenameEntryTransformerBench | bench_transform_10k_rows | 1 | 3 | 127.405mb +0.00% | 68.963ms -4.94% | ±0.90% +13.50% |
+-----------------------------+--------------------------+------+-----+------------------+-----------------+----------------+
Loaders+--------------------+----------------+------+-----+------------------+------------------+-----------------+
| benchmark | subject | revs | its | mem_peak | mode | rstdev |
+--------------------+----------------+------+-----+------------------+------------------+-----------------+
| CSVLoaderBench | bench_load_10k | 1 | 3 | 64.050mb +0.01% | 101.610ms -4.57% | ±1.05% +153.38% |
| JsonLoaderBench | bench_load_10k | 1 | 3 | 84.432mb +0.20% | 98.049ms -2.96% | ±0.57% -29.14% |
| ParquetLoaderBench | bench_load_10k | 1 | 3 | 161.284mb +0.01% | 20.982s -2.01% | ±0.34% -30.79% |
| TextLoaderBench | bench_load_10k | 1 | 3 | 18.138mb +0.04% | 31.429ms -0.90% | ±1.20% +13.69% |
+--------------------+----------------+------+-----+------------------+------------------+-----------------+
Building Blocks+-------------------+----------------------------+------+-----+------------------+------------------+-----------------+
| benchmark | subject | revs | its | mem_peak | mode | rstdev |
+-------------------+----------------------------+------+-----+------------------+------------------+-----------------+
| TypeDetectorBench | bench_type_detector | 1 | 3 | 43.920mb +0.02% | 374.932ms +2.31% | ±0.34% -2.64% |
| TypeDetectorBench | bench_type_detector | 1 | 3 | 11.730mb +0.06% | 75.612ms +2.15% | ±0.79% +12.10% |
| EntryFactoryBench | bench_entry_factory | 1 | 3 | 106.034mb +0.01% | 540.050ms +0.84% | ±1.57% +178.20% |
| EntryFactoryBench | bench_entry_factory | 1 | 3 | 55.224mb +0.01% | 267.383ms +0.32% | ±1.04% +19.08% |
| EntryFactoryBench | bench_entry_factory | 1 | 3 | 14.746mb +0.05% | 58.779ms +0.35% | ±1.46% +63.98% |
| RowsBench | bench_chunk_10_on_10k | 2 | 3 | 97.067mb +0.01% | 3.799ms -7.45% | ±2.22% -8.65% |
| RowsBench | bench_diff_left_1k_on_10k | 2 | 3 | 114.423mb +0.01% | 187.674ms +0.93% | ±0.31% -72.57% |
| RowsBench | bench_diff_right_1k_on_10k | 2 | 3 | 97.143mb +0.01% | 19.064ms +0.41% | ±0.54% -31.93% |
| RowsBench | bench_drop_1k_on_10k | 2 | 3 | 97.942mb +0.01% | 1.796ms -8.55% | ±1.59% -50.84% |
| RowsBench | bench_drop_right_1k_on_10k | 2 | 3 | 97.942mb +0.01% | 1.922ms -23.26% | ±1.57% -50.82% |
| RowsBench | bench_entries_on_10k | 2 | 3 | 96.103mb +0.01% | 5.572ms -4.79% | ±0.90% -59.17% |
| RowsBench | bench_filter_on_10k | 2 | 3 | 96.632mb +0.01% | 16.398ms -8.92% | ±0.26% -92.13% |
| RowsBench | bench_find_on_10k | 2 | 3 | 96.632mb +0.01% | 16.680ms -4.71% | ±1.50% +36.20% |
| RowsBench | bench_find_one_on_10k | 10 | 3 | 95.323mb +0.01% | 1.806μs -9.70% | ±2.57% +0.00% |
| RowsBench | bench_first_on_10k | 10 | 3 | 95.323mb +0.01% | 0.500μs 0.00% | ±0.00% 0.00% |
| RowsBench | bench_flat_map_on_1k | 2 | 3 | 104.541mb +0.01% | 16.530ms -2.87% | ±3.28% +137.14% |
| RowsBench | bench_map_on_10k | 2 | 3 | 134.608mb +0.00% | 75.053ms -5.16% | ±1.73% +43.85% |
| RowsBench | bench_merge_1k_on_10k | 2 | 3 | 97.151mb +0.01% | 1.964ms -6.58% | ±0.94% -73.03% |
| RowsBench | bench_partition_by_on_10k | 2 | 3 | 100.522mb +0.01% | 63.558ms -3.71% | ±0.98% +194.10% |
| RowsBench | bench_remove_on_10k | 2 | 3 | 98.204mb +0.01% | 4.202ms -6.95% | ±2.88% +28.38% |
| RowsBench | bench_sort_asc_on_1k | 2 | 3 | 95.685mb +0.01% | 42.446ms +0.23% | ±1.09% -43.15% |
| RowsBench | bench_sort_by_on_1k | 2 | 3 | 95.685mb +0.01% | 41.680ms -5.58% | ±1.05% -29.42% |
| RowsBench | bench_sort_desc_on_1k | 2 | 3 | 95.685mb +0.01% | 42.788ms -3.29% | ±1.71% -13.43% |
| RowsBench | bench_sort_entries_on_1k | 2 | 3 | 97.763mb +0.01% | 8.236ms -0.02% | ±0.88% -21.61% |
| RowsBench | bench_sort_on_1k | 2 | 3 | 95.513mb +0.01% | 29.987ms -1.27% | ±0.40% -82.44% |
| RowsBench | bench_take_1k_on_10k | 10 | 3 | 95.323mb +0.01% | 14.833μs -23.87% | ±2.75% -2.86% |
| RowsBench | bench_take_right_1k_on_10k | 10 | 3 | 95.323mb +0.01% | 15.983μs -29.48% | ±1.34% -35.17% |
| RowsBench | bench_unique_on_1k | 2 | 3 | 114.424mb +0.01% | 188.778ms +0.08% | ±0.48% -46.38% |
+-------------------+----------------------------+------+-----+------------------+------------------+-----------------+
|
Codecov ReportAttention: Patch coverage is
✅ All tests successful. No failed tests found. Additional details and impacted files@@ Coverage Diff @@
## 1.x #1577 +/- ##
==========================================
+ Coverage 83.13% 83.26% +0.13%
==========================================
Files 695 702 +7
Lines 18898 19012 +114
==========================================
+ Hits 15710 15831 +121
+ Misses 3188 3181 -7
🚀 New features to boost your workflow:
|
| $scale = 6; | ||
| } | ||
|
|
||
| var_dump(\json_encode(['scale' => $scale, 'left_scale' => $leftScale, 'right_scale' => $rightScale, 'left_value' => $leftValue, 'right_value' => $rightValue])); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
debug left in
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
(can be automated with https://github.com/spaze/phpstan-disallowed-calls)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
haha it happens to me way too often! Going to add that extension now 🤣
Thats why I release on a schedule, to give myself time to fix those kind of mistakes 🙈
Change Log
Added
Fixed
Changed
Removed
Deprecated
Security
Description
In general all ScalarFunctions can return any value
mixed. This means that the result of scalar function needs to go throughEntryFactoryin order to create an entry.Scalar functions can take their values from:
This means that functions like plus/minus/divide/multiply/mod will always return some float|integer and they wont keep the precision/scale of the input float (if the input was an entry).
In order to fix that, scalar functions can now also return
ScalarResult(mixed $value, Type $type)which is pretty much just a wrapper around the value.This approach is way more flexible that TypedScalarFunction since some return types are dynamic and depends on input arguments, for example
ref('id')->cast(type_int()).The return type of cast depends on the argument.
So now when ScalarResult is returned function can pass over more data, like for example type taken from the input.
But since not all values must come from a entry, all math functions received another argument
ScalarFunction|int $scale = 0which can be used to specify how precise that operation needs to be. Under the hood, scale is passed to the instance ofCalculator.