Suppress a health check¶
Hegel watches each run for signs that a test is set up in a way that will make
it slow or ineffective, and fails the run with a
FailedHealthCheck when it sees one. For example, a test that
throws away almost every input it generates:
@hegel.test
def test_multiples_of_1000(tc):
n = tc.draw(gs.integers())
tc.assume(n % 1000 == 0)
...
fails with a FailedHealthCheck whose message reads:
FailedHealthCheck: FilterTooMuch — it looks like this test is filtering out too
many inputs. 50 inputs were rejected by assume() or a filter while only 1 valid
inputs were generated. If this is expected, suppress the FilterTooMuch health
check with the suppress_health_check setting.
The failed health check is HealthCheck.filter_too_much; suppress it with
suppress_health_check=[HealthCheck.filter_too_much].
The exception’s health_check attribute is the HealthCheck
that failed, here HealthCheck.filter_too_much.
The best fix is usually to change the test, so that it generates the inputs it
wants directly. Here, gs.integers().map(lambda n: n * 1000) generates only
multiples of 1000. But when the behaviour is expected, suppress the health
check with the suppress_health_check setting:
from hegel import HealthCheck
@hegel.test(suppress_health_check=[HealthCheck.filter_too_much])
def test_multiples_of_1000(tc):
...
The setting takes a collection of HealthCheck members, or their
names (suppress_health_check=["filter_too_much"]).
suppress_health_check=HealthCheck.all() suppresses every health check.
To suppress a health check for every test, set it in a profile in
hegel.toml:
[profiles.development]
suppress_health_check = ["too_slow"]
The health checks¶
filter_too_muchToo many test cases are being discarded, by
assume()orfilter().too_slowThe first test cases are taking too long: more than 30 seconds in all before ten valid ones, counting the test body and its generators.
test_cases_too_largeThe test cases being generated are too large: they make too many choices. Suppressing this check also removes the limit on how many choices a test case can make.
large_initial_test_caseThe smallest input the test can generate is already very large, which usually means a generator has a large minimum size.