Skip to content

v0.9.0

Choose a tag to compare

View all tags
@jamesmckevitt jamesmckevitt released this 18 Sep 21:17
· 92 commits to master since this release
v0.9.0
b8d6143
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

New

  • simulation.noise: False turns the detector noise off, giving the signal the instrument would measure on average. (#75)
  • The 13 science cases of the Concept Study Report ship as one table, and eclipse-science-cases writes a uniform-intensity configuration for any case and line. (#85)
  • A fitting: block without components configures the single-Gaussian fit; every component's intensity, velocity and width are stored under its name; failed fits are counted; bessel_correction, save_iterations and component= in the analysis. (#83)
  • EIS uses its calibrated effective area curves: calibration chooses ground, dz2013, warren2014 or dz2025, with a date for the in-flight ones. (#59)
  • --hdf5-dbase-root chooses the CHIANTI database for one synthesis. (#55)
  • Maps carry an observation date (date_obs, now required) and a stated observer. (#79)
  • Configurations are checked: unknown keys, half-specified pinholes and uneven velocity or wavelength grids are refused with a message saying what to fix. (#76, #77, #78)
  • A pinhole can be placed at a chosen wavelength. (#58)

Documentation

A tutorial reproduces the paper's intensity and velocity precision maps end to end from the public simulation data, with the real run's timings. (#71, #87)

Infrastructure

Releases publish to PyPI automatically, the tests run in CI on every pull request, and the photon budget is checked against the standard radiometric equation. (#48, #60)

Changes to output

  • EIS runs use the calibrated effective area instead of a constant. (#59)
  • The spatial PSF sees the Sun continuing beyond the raster, so the outermost slit rows are slightly brighter. (#80)
  • --downsample scales the voxel sizes once; downsampled synthesis files from earlier versions need regenerating. (#82)
  • Line-of-sight velocity is positive away from the observer, who is above the box for a view along z and at +x for a view along x. (#86)
  • Pinhole light that falls outside the detector is no longer redistributed onto it, so pinhole contributions are lower, most for small pinholes. (#58)
  • Tied components are tied by velocity rather than wavelength separation, and a single-Gaussian fit that did not converge in scipy's default number of iterations gets 1000. (#57)
  • Fits that fail are left out of the statistics, and each component uses its own rest wavelength. (#83)
  • Cubes are stored (y, x, wavelength), so a seeded run lands its random draws on different pixels than before and earlier pickles are refused. (#61)
Assets 2
Loading