✨ Add plot_doas function that only plots DOAS related information

[s-weigand]

This PR replaces the current plot_doas function which was meant as an augmented plot_overview and is broken on main with a more focused (3 plots) version that only plots DOAS-related information.
In addition, it adds a custom plot tick calculation function for values that are in the unit of Pi.

Original plots by @ism200 in the publication of the 4TT case study

Plots using plot_doas and the pyglotaran results

Note: The slight variations are due to a slightly different model using DOAS to fit parts of the coherent artifact for better fit results and the flipped phase for osc1 is due to using np.unwrap in pyglotaran in contrast to mod 2*Pi in the original plot

As for adding \overline{\nu}... as text labels to the plots this should be done via a helper function since it would need a latex environment at the user side installed, which we should not force on them to have by default.

Pseudo plot code e.g. for beta_doas

from pyglotaran_extras.plotting.plot_doas import plot_doas
from glotaran.io import load_result

result = load_result("doas_case_study/result.yaml")

fig, axes = plot_doas(
    result,
    damped_oscillation=["osc1", "osc2", "osc5"],
    time_range=(-0.8, 0.8),
    spectral=550,
    center_λ=550,
    # oscillation_type="sin"
    # normalize=False
)

Change summary

• ✨ Added calculate_ticks_in_units_of_pi utility function
• ✨ Added plot_doas function
• 👌 Allow None as input to extract_irf_location to deactivate shifting
• 👌 Remove center_λ as argument to plot_doas and always shift at spectral
• 👌 Use main_irf_nr=None to optout of shifting by the IRF
• 👌 Changed default of normalize in plot_doas to False

Checklist

• ✔️ Passing the tests (mandatory for all PR's)
• 🧪 Adds new tests for the feature (mandatory for ✨ feature and 🩹 bug fix PR's)

[sourcery-ai]

Sourcery Code Quality Report

❌  Merging this PR will decrease code quality in the affected files by 1.87%.

Quality metrics	Before	After	Change
Complexity	3.31 ⭐	3.13 ⭐	-0.18 👍
Method Length	96.33 🙂	96.30 🙂	-0.03 👍
Working memory	7.81 🙂	10.08 😞	2.27 👎
Quality	69.34% 🙂	67.47% 🙂	-1.87% 👎

Other metrics	Before	After	Change
Lines	547	672	125

Changed files	Quality Before	Quality After	Quality Change
pyglotaran_extras/__init__.py	84.86% ⭐	83.57% ⭐	-1.29% 👎
pyglotaran_extras/plotting/plot_doas.py	45.89% 😞	33.35% 😞	-12.54% 👎
pyglotaran_extras/plotting/utils.py	75.07% ⭐	75.31% ⭐	0.24% 👍
tests/plotting/test_utils.py	82.08% ⭐	80.22% ⭐	-1.86% 👎

Here are some functions in these files that still need a tune-up:

File	Function	Complexity	Length	Working Memory	Quality	Recommendation
pyglotaran_extras/plotting/plot_doas.py	plot_doas	6 ⭐	380 ⛔	22 ⛔	32.05% 😞	Try splitting into smaller methods. Extract out complex expressions
pyglotaran_extras/plotting/utils.py	select_plot_wavelengths	3 ⭐	136 😞	11 😞	59.58% 🙂	Try splitting into smaller methods. Extract out complex expressions
tests/plotting/test_utils.py	test_abs_max	0 ⭐	136 😞	5 ⭐	76.11% ⭐	Try splitting into smaller methods

Legend and Explanation

The emojis denote the absolute quality of the code:

• ⭐ excellent
• 🙂 good
• 😞 poor
• ⛔ very poor

The 👍 and 👎 indicate whether the quality has improved or gotten worse with this pull request.

---

Please see our documentation here for details on how these metrics are calculated.

We are actively working on this report - lots more documentation and extra metrics to come!

Help us improve this quality report!

[github-actions]

👈 Launch a binder notebook on branch s-weigand/pyglotaran-extras/plot-doas

[codecov]

Codecov Report

Base: 43.25% // Head: 44.97% // Increases project coverage by +1.72% 🎉

Coverage data is based on head (3e96084) compared to base (c1385c9).
Patch coverage: 30.43% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #135      +/-   ##
==========================================
+ Coverage   43.25%   44.97%   +1.72%     
==========================================
  Files          29       29              
  Lines         904      916      +12     
  Branches      133      133              
==========================================
+ Hits          391      412      +21     
+ Misses        506      497       -9     
  Partials        7        7              

Impacted Files	Coverage Δ
pyglotaran_extras/plotting/plot_doas.py	31.81% <18.91%> (+31.81%)	⬆️
pyglotaran_extras/plotting/utils.py	29.35% <75.00%> (+3.61%)	⬆️
pyglotaran_extras/__init__.py	100.00% <100.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[s-weigand]

@ism200 @jsnel I'm not 100% sure if we want to keep the slitting between spectral (wavelength we select the oscillations data from) and center_λ (wavelength we select the IRF to shift the oscillations).
Maybe using a default of None for center_λ and replacing it with the value of spectral in that case would be the best of two worlds, but in that case it would kinda hide the dispersion from the user if they were only looking at the DOAS with our plot functions.

[jsnel]

Much improved!

[s-weigand]

As requested by @ism200 I streamlined plot_doas

• 👌 Allow None as input to extract_irf_location to deactivate shifting
• 👌 Remove center_λ as argument to plot_doas and always shift at spectral
• 👌 Use main_irf_nr=None to optout of shifting by the IRF
• 👌 Changed default of normalize in plot_doas to False

[jsnel]

For posterity I record that @ism200 reviewed these changes and acknowledges the (subtle) difference between the two plots that are shown, as captured by this comment:

Note: The slight variations are due to a slightly different model using DOAS to fit parts of the coherent artifact for better fit results and the flipped phase for osc1 is due to using np.unwrap in pyglotaran in contrast to mod 2*Pi in the original plot

He commented that the unwrap method is in fact better, so from my perspective this is good to be merged in.

If you wish to acknowledge this @ism200, please press the Files changed button on top and review as 'ok'.

[s-weigand]

New plot functionality

Pseudo plot code e.g. for ex_doas_beta

from pyglotaran_extras.plotting.plot_doas import plot_doas
from glotaran.io import load_result

result = load_result("doas_case_study/result.yaml")

fig, axes = plot_doas(
    result,
    damped_oscillation=["osc1", "osc2", "osc5"],
    time_range=(-0.8, 0.8),
    spectral=550,
    # oscillation_type="sin"
    # normalize=True
)

for vline_pos in [412]:
    axes[1].axvline(vline_pos, color="k", linewidth=1)
    axes[2].axvline(vline_pos, color="k", linewidth=1)

[ism200]

this looks all very nice now. thanks a lot!

[ism200]

thanks a lot !