v0.3.0

Highlights

Adds ctis.instruments.OptikaInstrument, a CTIS instrument whose forward model is an optika LinearSystem, and moves the instrument interface to an electron-based measurement model.

New

• ctis.instruments.OptikaInstrument — forward/backward model backed by an optika.systems.LinearSystem, supplying channel-aware distortion, effective area, and vignetting via cached regridding weights.
• IdealInstrument gains quantum_yield and read_noise; image returns electrons, integrates once per readout, and adds read noise once.

Changed (breaking)

• image gains an uncertainty flag returning a NormalUncertainScalarArray; the standalone uncertainty property is removed. merit.mean_chi_squared now takes the uncertainty as an argument.
• backproject gains a unit argument to express the result in photon or energy units.
• regrid and the instrument volume weighting accept Doppler grids (normalized via .spectral_positional).

Requires

• optika ~= 2.1
• named-arrays ~= 2.4

v0.2.0

What's Changed

• Added a percentile_radiance argument to AbstractInversionResult.plot_moments() to exclude dim pixels from the histogram. by @roytsmart in #15
• Modified AbstractInversionResult.plot_moment() to print Pearson's correlation coefficient. by @roytsmart in #16
• Update named-arrays to v2.0. by @roytsmart in #17

Full Changelog: v0.1.1...v0.2.0

v0.1.1

What's Changed

• Removed logic that failed the MART inversion if the merit function was increasing. by @roytsmart in #13
• Added a missing axis argument to AbstractInversionResult.plot_moments(). by @roytsmart in #14

Full Changelog: v0.1.0...v0.1.1

v0.1.0

What's Changed

• Added a discussion by @kankelborg about the analogy between Richardson-Lucy deconvolution and MART. by @roytsmart in #3
• Fix description in pyproject.toml. by @roytsmart in #4
• Added ctis.scenes.random_gaussians(), a test pattern developed by Amy Winebarger. by @roytsmart in #5
• Added ctis.instruments module. by @roytsmart in #1
• Fixed a bug where AbstractLinearInstrument.weights_transpose() was not conserving flux. by @roytsmart in #7
• Added axis_channel and num_channel properties to AbstractInstrument. by @roytsmart in #8
• Added an uncertainty property to AbstractInstrument. by @roytsmart in #9
• Modified the image() and backproject() methods of AbstractInstrument to accept instances of na.FunctionArray. by @roytsmart in #11
• Added ctis.inverters.MartInverter, an implementation of the Multiplicative Algebraic Reconstruction Technique (MART). by @roytsmart in #10

New Contributors

• @roytsmart made their first contribution in #3

Full Changelog: v0.0.0...v0.1.0

v0.0.0

Initial empty release.

Full Changelog: https://github.com/sun-data/ctis/commits/v0.0.0