Closed jeremykubica closed 2 weeks ago
Before [18536db2] | After [a72a6008] | Ratio | Benchmark (Parameter) |
---|---|---|---|
751±800μs | 772±500μs | 1.03 | benchmarks.TimeSuite.time_fnu_to_flam |
4.44±0.3ms | 4.52±0.5ms | 1.02 | benchmarks.TimeSuite.time_chained_evaluate |
8.79±0.8ms | 8.92±1ms | 1.02 | benchmarks.TimeSuite.time_evaluate_salt3_model |
27.1±0.4μs | 27.6±0.2μs | 1.02 | benchmarks.TimeSuite.time_make_new_salt3_model |
15.4±0.1μs | 15.7±0.2μs | 1.02 | benchmarks.TimeSuite.time_sample_x1_from_hostmass |
4.99±0.3ms | 4.97±0.1ms | 1 | benchmarks.TimeSuite.time_evaluate_salt3_passbands |
1.43±0.06ms | 1.42±0.1ms | 0.99 | benchmarks.TimeSuite.time_apply_passbands |
8.97±0.04ms | 8.89±0.02ms | 0.99 | benchmarks.TimeSuite.time_load_passbands |
1.37±0.02s | 1.36±0.01s | 0.99 | benchmarks.TimeSuite.time_make_x1_from_hostmass |
123±2μs | 121±3μs | 0.98 | benchmarks.TimeSuite.time_sample_x0_from_distmod |
Click here to view all benchmarks.
Building off the pzflow sampling request (PR #181), add the ability to pre-sample data into an AstroPy table or Pandas DataFrame and then retrieve it during sampling.
This could look like: