XRT DEM Iterative Solver Module - #367
Conversation
|
Ensure the function checks for minimum required inputs (filters, observed intensities, etc.) and validates that all observation values are non-zero before proceeding. |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #367 +/- ##
===========================================
- Coverage 87.97% 36.96% -51.02%
===========================================
Files 26 40 +14
Lines 1855 4415 +2560
Branches 110 338 +228
===========================================
Hits 1632 1632
- Misses 180 2740 +2560
Partials 43 43 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
2041e0a to
bd4687b
Compare
…ng at each temperature
| monte_carlo_runs=0, | ||
| ) | ||
|
|
||
| x.validate_inputs() |
There was a problem hiding this comment.
Should validate_inputs() be changed to _validate_inputs()? Is it something that needs to be user-facing?
| minimum_bound_temperature=5.5, | ||
| maximum_bound_temperature=7.5, | ||
| logarithmic_temperature_step_size=0.1, |
There was a problem hiding this comment.
| minimum_bound_temperature=5.5, | |
| maximum_bound_temperature=7.5, | |
| logarithmic_temperature_step_size=0.1, | |
| **temperature_kwargs, |
And define temperature_kwargs = {"minimum_bound_temperature": 5.5, ...} outside of the function, so that we can use keyword argument unpacking here?
|
|
||
| x.create_logT_grid() | ||
|
|
||
| assert (x.logT[0], x.logT[-1], x.dlogT) == pytest.approx((5.5, 7.5, 0.1)) |
There was a problem hiding this comment.
| assert (x.logT[0], x.logT[-1], x.dlogT) == pytest.approx((5.5, 7.5, 0.1)) | |
| np.testing.assert_allclose( | |
| [x.logT[0], x.logT[-1], x.dlogT], | |
| [5.5, 7.5, 0.1], | |
| ) |
| dem.validate_inputs() | ||
|
|
||
|
|
||
| def test_create_logT_grid(): |
There was a problem hiding this comment.
This would be a good use case for pytest fixtures to avoid having repeated assertions, but it's probably not worth the effort changing here.
| x.create_logT_grid() | ||
| x._interpolate_responses_to_grid() | ||
|
|
||
| dem, modeled, chi2, result = x._solve_single_dem( |
There was a problem hiding this comment.
🎗️ I'm wondering if this would also be useful to think about adding upstream to the interface in sunkit-dem, especially if many DEM solvers are iterative. 🤔
| temperature_responses=responses, | ||
| ) | ||
| x.solve() | ||
| x.plot_dem() |
There was a problem hiding this comment.
🎗️ Is there a way to do something like a regression test for figures?
| def test_init_rejects_negative_max_iterations(): | ||
| """max_iterations=-100 must raise ValueError. | ||
| Expected: ValueError because max_iterations must be a positive integer. | ||
| """ | ||
| filters = ["Al-poly", "Ti-poly"] | ||
| intensities = np.array([500.0, 800.0]) | ||
| responses = generate_temperature_responses(filters, "2012-10-27T00:00:00") | ||
|
|
||
| with pytest.raises(ValueError, match="max_iterations must be a positive integer"): | ||
| XRTDEMIterative(filters, intensities, responses, max_iterations=-100) | ||
|
|
||
|
|
||
| def test_init_rejects_zero_normalization_factor(): | ||
| """normalization_factor=0 must raise ValueError. | ||
| Expected: ValueError because normalization_factor must be positive. | ||
| """ | ||
| filters = ["Al-poly", "Ti-poly"] | ||
| intensities = np.array([500.0, 800.0]) | ||
| responses = generate_temperature_responses(filters, "2012-10-27T00:00:00") | ||
|
|
||
| with pytest.raises(ValueError, match="normalization_factor must be a positive"): | ||
| XRTDEMIterative(filters, intensities, responses, normalization_factor=0) |
There was a problem hiding this comment.
We might be able to use @pytest.mark.parametrize for some of these. 🤔
| filters = ["Al-poly", "Ti-poly", "Be-thin"] | ||
| intensities = np.array([100.0, 200.0, 300.0], dtype=float) |
There was a problem hiding this comment.
Some of these be defined outside of the test functions to reduce duplication.
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Removing space. Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Removing space pt2. Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Nice touch. Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
Co-authored-by: Nick Murphy <namurphy@users.noreply.github.com>
2 participants
Summary
This PR introduces
XRTDEMIterative, a Python implementation of the IDLroutine
xrt_dem_iterative2.profor estimating Differential EmissionMeasures (DEMs) from Hinode/XRT multi-filter observations.
What's Included
Core Solver (
dem_solver.py)XRTDEMIterativeclass with full DEM solving pipelinefitting via
lmfitmax(0.03 * intensity, 2 DN/s/pix), matching IDL default behaviormonte_carlo_runsparameterintensities, missing errors)
summary()method for diagnostic outputPlotting (
dem_plotting.py)plot_dem()— base DEM solution visualizationplot_dem_mc()— base DEM + Monte Carlo ensemble overlayTests
response interpolation, spline system, residuals, solver, and
Monte Carlo behavior
Documentation
dem_overview.rst— introduction to DEM analysis with XRTpyxrtpy.xrt_dem_iterativedocs/changelog/0.6.0.rstValidation -
xrt_dem_iterative2.proacross multiplesolar observation cases using Monte Carlo comparison