swh:1:snp:e65ec12a3dc489100df2dfc1b0eab4915109f830
Tip revision: d6748f9a1e915d476d4fadb28bb5dd6d378b2286 authored by Jesper Nielsen on 29 September 2022, 15:30:57 UTC
Merge pull request #1993 from GPflow/develop
Merge pull request #1993 from GPflow/develop
Tip revision: d6748f9
test_benchmark_api.py
# Copyright 2022 The GPflow Contributors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from typing import Collection
from unittest.mock import Mock
import pandas as pd
import pytest
import benchmark.datasets as ds
import benchmark.models as md
import benchmark.plotters as pl
from benchmark.benchmark_api import BenchmarkSet, BenchmarkSuite, BenchmarkTask
from benchmark.grouping import GroupingKey as GK
from benchmark.grouping import GroupingSpec
@pytest.mark.parametrize(
"benchmark_set,expected_tasks",
[
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
ds.tiny_sine,
],
models=[
md.gpr,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False],
do_optimise=[False],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
[
BenchmarkTask(
dataset_name="tiny_linear",
model_name="gpr",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
BenchmarkTask(
dataset_name="tiny_sine",
model_name="gpr",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
],
),
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
],
models=[
md.gpr,
md.svgp,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False],
do_optimise=[False],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
[
BenchmarkTask(
dataset_name="tiny_linear",
model_name="gpr",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
BenchmarkTask(
dataset_name="tiny_linear",
model_name="svgp",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
],
),
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
],
models=[
md.gpr,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False, True],
do_optimise=[False],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
[
BenchmarkTask(
dataset_name="tiny_linear",
model_name="gpr",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
BenchmarkTask(
dataset_name="tiny_linear",
model_name="gpr",
do_compile=True,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
],
),
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
],
models=[
md.gpr,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False],
do_optimise=[False, True],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
[
BenchmarkTask(
dataset_name="tiny_linear",
model_name="gpr",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
BenchmarkTask(
dataset_name="tiny_linear",
model_name="gpr",
do_compile=False,
do_optimise=True,
do_predict=True,
do_posterior=True,
repetitions=2,
),
],
),
],
)
def test_benchmark_set__get_tasks(
benchmark_set: BenchmarkSet, expected_tasks: Collection[BenchmarkTask]
) -> None:
assert expected_tasks == benchmark_set.get_tasks()
@pytest.mark.parametrize(
"benchmark_set,expected_metrics",
[
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
ds.tiny_sine,
],
models=[
md.gpr,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False],
do_optimise=[False],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
pd.DataFrame(
[
(0, "tiny_linear", "gpr", False, False, 0),
(1, "tiny_sine", "gpr", False, False, 0),
(5, "tiny_linear", "gpr", False, False, 1),
],
columns=["id", "dataset", "model", "do_compile", "do_optimise", "repetition"],
),
),
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
],
models=[
md.gpr,
md.svgp,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False],
do_optimise=[False],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
pd.DataFrame(
[
(0, "tiny_linear", "gpr", False, False, 0),
(2, "tiny_linear", "svgp", False, False, 0),
(5, "tiny_linear", "gpr", False, False, 1),
],
columns=["id", "dataset", "model", "do_compile", "do_optimise", "repetition"],
),
),
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
],
models=[
md.gpr,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False, True],
do_optimise=[False],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
pd.DataFrame(
[
(0, "tiny_linear", "gpr", False, False, 0),
(3, "tiny_linear", "gpr", True, False, 0),
(5, "tiny_linear", "gpr", False, False, 1),
],
columns=["id", "dataset", "model", "do_compile", "do_optimise", "repetition"],
),
),
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
],
models=[
md.gpr,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False],
do_optimise=[False, True],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=2,
),
pd.DataFrame(
[
(0, "tiny_linear", "gpr", False, False, 0),
(4, "tiny_linear", "gpr", False, True, 0),
(5, "tiny_linear", "gpr", False, False, 1),
],
columns=["id", "dataset", "model", "do_compile", "do_optimise", "repetition"],
),
),
(
BenchmarkSet(
name="metrics",
datasets=[
ds.tiny_linear,
],
models=[
md.gpr,
],
plots=[
pl.metrics_box_plot,
],
do_compile=[False],
do_optimise=[False],
do_predict=True,
do_posterior=True,
file_by=GroupingSpec((GK.DATASET,), minimise=False),
row_by=GroupingSpec((GK.METRIC,), minimise=False),
column_by=GroupingSpec((GK.PLOTTER,), minimise=False),
line_by=None,
repetitions=5,
),
pd.DataFrame(
[
(0, "tiny_linear", "gpr", False, False, 0),
(5, "tiny_linear", "gpr", False, False, 1),
(6, "tiny_linear", "gpr", False, False, 2),
(7, "tiny_linear", "gpr", False, False, 3),
],
columns=["id", "dataset", "model", "do_compile", "do_optimise", "repetition"],
),
),
],
)
def test_benchmark_set__filter_metrics(
benchmark_set: BenchmarkSet,
expected_metrics: pd.DataFrame,
) -> None:
metrics = pd.DataFrame(
[
(0, "tiny_linear", "gpr", False, False, 0),
(1, "tiny_sine", "gpr", False, False, 0),
(2, "tiny_linear", "svgp", False, False, 0),
(3, "tiny_linear", "gpr", True, False, 0),
(4, "tiny_linear", "gpr", False, True, 0),
(5, "tiny_linear", "gpr", False, False, 1),
(6, "tiny_linear", "gpr", False, False, 2),
(7, "tiny_linear", "gpr", False, False, 3),
],
columns=["id", "dataset", "model", "do_compile", "do_optimise", "repetition"],
)
after = benchmark_set.filter_metrics(metrics).reset_index(drop=True)
pd.testing.assert_frame_equal(expected_metrics, after)
def test_benchmark_suite__get_tasks() -> None:
set1 = Mock(BenchmarkSet)
set1.name = "set1"
set1.get_tasks.return_value = [
BenchmarkTask(
dataset_name="ds",
model_name="md",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=False,
repetitions=2,
),
BenchmarkTask(
dataset_name="ds1",
model_name="md1",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
]
set2 = Mock(BenchmarkSet)
set2.name = "set2"
set2.get_tasks.return_value = [
BenchmarkTask(
dataset_name="ds",
model_name="md",
do_compile=False,
do_optimise=False,
do_predict=False,
do_posterior=True,
repetitions=2,
),
BenchmarkTask(
dataset_name="ds2",
model_name="md1",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
]
set3 = Mock(BenchmarkSet)
set3.name = "set3"
set3.get_tasks.return_value = [
BenchmarkTask(
dataset_name="ds",
model_name="md",
do_compile=False,
do_optimise=False,
do_predict=False,
do_posterior=False,
repetitions=5,
),
BenchmarkTask(
dataset_name="ds1",
model_name="md3",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
]
suite = BenchmarkSuite(
name="test", description="Suite used in a test.", sets=[set1, set2, set3]
)
assert [
BenchmarkTask(
dataset_name="ds",
model_name="md",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=5,
),
BenchmarkTask(
dataset_name="ds1",
model_name="md1",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
BenchmarkTask(
dataset_name="ds2",
model_name="md1",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
BenchmarkTask(
dataset_name="ds1",
model_name="md3",
do_compile=False,
do_optimise=False,
do_predict=True,
do_posterior=True,
repetitions=2,
),
] == suite.get_tasks()