-
Notifications
You must be signed in to change notification settings - Fork 101
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
Tools for dealing with limited precision #1392
Conversation
In particular, the test suite should refuse to run at all, with a message about precision. I'd appreciate if someone could comment out the line in |
Longer-term, we could go through and flag which tests require full precision and disable running of those tests on low-precision machines. |
I tried this on my M1 Mac and it works as advertised. |
src/pint/simulation.py
Outdated
) | ||
|
||
|
||
def update_fake_toa_clock(ts, model, include_bipm=False, include_gps=True): | ||
"""Update the clock settings (corrections, etc) for fake TOAs | ||
"""Update the clock s:qettings (corrections, etc) for fake TOAs |
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.
Typo
Codecov Report
@@ Coverage Diff @@
## master #1392 +/- ##
==========================================
- Coverage 62.25% 62.24% -0.01%
==========================================
Files 89 89
Lines 20267 20278 +11
Branches 3653 3657 +4
==========================================
+ Hits 12617 12623 +6
- Misses 6860 6862 +2
- Partials 790 793 +3
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. |
As we see in #1389, users can miss or ignore the "limited precision" warning when trying to run PINT with 64-bit longdouble values. This PR tries to make PINT run less badly and signal more loudly in this setting.
pint.simulation
will make fake TOAs but only good to a few microseconds.require_longdouble_precision()
to do the test and get an informative exception. I haven't identified any such functions yet (do Fitters work okay, for some values of "okay"?)I would appreciate additional testing in a 64-bit-only environment (M1 macs, Windows MSVC).