Addressing more test warnings

[Yash-10]

Continuing work from #1137. Thanks @ishanSrt !

Some analysis

• The dubious year warnings (coming from test_propagate_long_times_keeps_geometry) due to the dates being far into the future seem to be legitimate, since that test propagates an orbit for 100 years and hence the resultant orbit is mostly going to be out of the non-dubious range defined by erfa.
• I thought it might be a good choice to just filter the warning since after looking at issue 256, I wondered whether it would be good to still check for long propagations like 100 years. Moreover, reducing the propagation time while still keeping it long enough, say, 60 years didn't seem to solve it.
• For the second set of erfa warnings, eg:

ERFA function "epv00" yielded 8 of "warning: date outsidethe range 1900-2100 AD"

Any date after about the year 1935 yields that warning. Here, it might be a good choice to just choose a "safe" date value to prevent these warnings pop up.

Thanks, and would love to get any suggestions!

---

The coverage decrease needs to be addressed and I am working on that...

[codecov]

Codecov Report

Merging #1235 (7b0aa05) into main (112cba6) will increase coverage by 0.40%.
The diff coverage is 88.88%.

❗ Current head 7b0aa05 differs from pull request most recent head 0be91a0. Consider uploading reports for the commit 0be91a0 to get more accurate results

@@            Coverage Diff             @@
##             main    #1235      +/-   ##
==========================================
+ Coverage   90.82%   91.23%   +0.40%     
==========================================
  Files          78       78              
  Lines        4121     4106      -15     
  Branches      363      360       -3     
==========================================
+ Hits         3743     3746       +3     
+ Misses        288      270      -18     
  Partials       90       90              

Impacted Files	Coverage Δ
src/poliastro/czml/extract_czml.py	88.09% <0.00%> (ø)
src/poliastro/twobody/orbit.py	82.83% <ø> (+3.18%)	⬆️
src/poliastro/plotting/misc.py	94.73% <100.00%> (+0.98%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 112cba6...0be91a0. Read the comment docs.

[Yash-10]

Points and questions regarding coverage:

• The test test_misc.py is modified such that no warnings are raised up. Doing this has decreased coverage for misc.py since it now doesn't test for the case when no epoch argument is passed, i.e. the default epoch of J2000 is used which, as per the above analysis, is bound to generate "warning: date outsidethe range 1900-2100 AD".

Since testing for the case when no epoch argument is passed might be more important than handling warnings, it might make sense to just suppress the warning if no better solution can be reached.

• The reason for ignoring the dubious year warnings, again from the PR description might be a good choice since testing for long propagations might be important.

Question
Since Orbit.pqw() is now deprecated and is going to be removed (as per the warning message) and that tests for it are also removed, would it be fine to remove Orbit.pqw completely? I think this might stabilize the coverage because currently the tests for pqw are removed but it is still in the source code. Just wanted to clarify this before making any changes :)

Thanks!

[Yash-10]

@astrojuanlu Could you please guide me on the above and let me know your thoughts?

Thanks!

[astrojuanlu]

Hello @Yash-10! I'm about to review this PR. Would you please rebase on top of main?

[astrojuanlu]

Did a thorough review, thanks for your patience!

[astrojuanlu]

This looks good to go! My only remaining question is: did we have to regenerate the images? The changes look like a 1 pixel displacement, and a diff on the tens or hundreds of bytes. Were image tests failing locally because of it? If not, I think we should remove the changes and squash the commit, to avoid increasing the size of the repository.

[astrojuanlu]

Merging & squashing after the tests pass 👍🏽

[astrojuanlu]

(Not sure the automerge will work because of the coverage check, but let's see)