Releases: sot/timbre
Version 1.2
This release makes the following changes:
- Adds the ability to add custom anchors to a composite simulation
- Adds a "safe" ACIS state (HRC)
- Fixes a spelling mistake
Version 1.1.1
Fix some docstring errors, remove some unused imports.
Version 1.1
This updates the load_model_specs() function to default to using a local chandra_models repository instead of a Github hosted chandra_models repository. A separate function, load_github_model_specs() was added so users can continue to use a Github hosted repository if needed. Both of these functions calculate and supply the md5 hash for each model, and supply the latest commit sha for the version returned. These values are necessary for better tracking of Timbre-produced results.
Version 1.0.1
This fixes a bug that prevented one from using a custom chandra_models directory to supply Xija models.
Version 1.0
This is the official 1.0 release, roughly two and a half years in the making! Thank you to everyone who helped test, scrutinize, criticize, plot, ponder, listen, and give me their thoughts on how to make sure this is a useful tool!
Version 0.13
This adds the min_dwell keyword option to the find_second_dwell method, as well as the run_state_pairs method which calls the former method. This option allows one to manually narrow the search range for a balance dwell.
Version 0.12
This updates the docs to use Jupyter notebooks to serve as the primary documentation tool. Docs and tests were also updated to reflect the latest set of models.
Version 0.11.1
This adds some plotting code necessary to produce the random schedule plot in the documentation_material.ipynb notebook.
Version 0.11
This overhauls the balance algorithm that generates a composite maximum dwell estimate.
- The composite maximum dwell estimate calculation is much more robust and is able to handle additional edge cases
- One test was updated to accommodate the balance algorithm changes.
- Content was added to the
documentation_material.ipynbnotebook.
Version 0.10.0
This release makes the following changes:
- Return data types for the
run_state_pairs()function no longer need to be explicitly identified. - Additional data is returned by
run_state_pairs()to aid in categorizing results. - Balance time refinement now uses linear interpolation instead of quadratic interpolation to avoid problems with edge cases.
- The
Balanceclass now specifiesn_dwells=30for calls tofind_second_dwell()andrun_state_pairs(). This helps to offset increased error from using linear interpolation for balance time refinement, but increases the run time by a factor of 2 or 3. - The prior multiprocessing example is removed.
- A new multiprocessing example is added which generates data required to determine composite dwell limits directly.
- There are also a number of minor bug fixes included in this release.