Closed MichaelHatherly closed 11 months ago
main | 2178c8d50f10ef... | t[main]/t[2178c8d50f10ef...] | |
---|---|---|---|
Quantity/creation/Quantity(x) | 3.4 ± 0 ns | 3.7 ± 0.1 ns | 0.919 |
Quantity/creation/Quantity(x, length=y) | 3.4 ± 0 ns | 3.6 ± 0.1 ns | 0.944 |
Quantity/with_numbers/*real | 3.1 ± 0.1 ns | 3 ± 0.1 ns | 1.03 |
Quantity/with_numbers/^int | 12 ± 4.1 ns | 11.8 ± 4 ns | 1.02 |
Quantity/with_numbers/^int * real | 12.2 ± 4.7 ns | 12.1 ± 4.8 ns | 1.01 |
Quantity/with_quantity/+y | 7.1 ± 0 ns | 7.1 ± 0.1 ns | 1 |
Quantity/with_quantity//y | 3.7 ± 0 ns | 3.7 ± 0 ns | 1 |
Quantity/with_self/dimension | 1.7 ± 0 ns | 1.7 ± 0 ns | 1 |
Quantity/with_self/inv | 3.5 ± 0.2 ns | 3.4 ± 0.2 ns | 1.03 |
Quantity/with_self/ustrip | 1.7 ± 0 ns | 1.7 ± 0 ns | 1 |
QuantityArray/broadcasting/multi_array_of_quantities | 0.233 ± 0.028 ms | 0.232 ± 0.03 ms | 1.01 |
QuantityArray/broadcasting/multi_normal_array | 0.0772 ± 0.0027 ms | 0.0769 ± 0.0026 ms | 1 |
QuantityArray/broadcasting/multi_quantity_array | 0.255 ± 0.0017 ms | 0.255 ± 0.0016 ms | 1 |
QuantityArray/broadcasting/x^2_array_of_quantities | 0.0639 ± 0.017 ms | 0.0645 ± 0.016 ms | 0.991 |
QuantityArray/broadcasting/x^2_normal_array | 12.5 ± 3.4 μs | 12.5 ± 3.7 μs | 1 |
QuantityArray/broadcasting/x^2_quantity_array | 13.3 ± 1.3 μs | 13.4 ± 1.2 μs | 0.993 |
QuantityArray/broadcasting/x^4_array_of_quantities | 0.15 ± 0.0075 ms | 0.148 ± 0.0066 ms | 1.01 |
QuantityArray/broadcasting/x^4_normal_array | 0.0719 ± 0.0029 ms | 0.0708 ± 0.003 ms | 1.02 |
QuantityArray/broadcasting/x^4_quantity_array | 0.108 ± 0.0015 ms | 0.107 ± 0.0026 ms | 1.01 |
time_to_load | 0.214 ± 0.00081 s | 0.216 ± 0.0031 s | 0.991 |
A plot of the benchmark results have been uploaded as an artifact to the workflow run for this PR. Go to "Actions"->"Benchmark a pull request"->[the most recent run]->"Artifacts" (at the bottom).
Looks good! The one suggestion I have is to mention these units and aliases in this line:
"Time in seconds. Available variants: `fs`, `ps`, `ns`, `μs` (/`us`), `ms`, `min`, `h` (/`hr`), `day`, `yr`, `kyr`, `Myr`, `Gyr`.",
Thanks for the addition 👍
Thanks Miles.
There are a couple of differences between the naming of units in
Unitful
and this package that makes it not quite a direct swap when changing between the two packages. This adds a couple of aliases for current units as well as a newwk
.The only one that we're really needing here is
wk
since we can deal with aliasing the others internally where needed, so I'm fine with dropping those if there are reasons that they weren't included already?How would you like tests written for these kinds of additions? If they are needed?