v0.3.0

Commits

• 1c80f90: Fix a code block in the mdbook. (Christopher H. Jordan)
• 8f6bb41: Tweak cargo compilation profiles. (Christopher H. Jordan)
• 2ed1164: Add new errors. (Christopher H. Jordan)
• 077c88a: Clippy lints and project tweaks. (Christopher H. Jordan)
• 5c55728: Allow benchmarks to run by making relevant types public. (Christopher H. Jordan)
• d15145c: Improvements to calibration code. (Christopher H. Jordan)
• 4021a47: Improvements to modelling code. (Christopher H. Jordan)
• 048a76c: Update dependencies. Fix clippy lints. (Christopher H. Jordan)
• 6e510e9: Source list improvements. (Christopher H. Jordan)
• b2001f7: vis_io -> io (Christopher H. Jordan)
• 475613e: Fix raw data timestamp bug. (Christopher H. Jordan)
• b31adb5: Fix raw data obs with all flagged tiles being rejected. (Christopher H. Jordan)
• 3e2378e: Overhaul coarse channel info. (Christopher H. Jordan)
• 6f44f82: Fix input data array positions being ignored. (Christopher H. Jordan)
• 48fe022: Add benchmarks for IO reading. (Christopher H. Jordan)
• a1f9c6d: Make uvfits reading more robust. (Christopher H. Jordan)
• 72c67a4: Make MS reading more robust. (Christopher H. Jordan)
• 737e459: Handle multiple combinations of polarisations. (Christopher H. Jordan)
• 6995d0b: Support input data with the alternative baseline layout. (Christopher H. Jordan)
• 9ceaea6: Improve plotting code. (Christopher H. Jordan)
• 4b5f742: Improve UX with a new expensive_op function. (Christopher H. Jordan)
• 56e78bd: CLI code overhaul. (Christopher H. Jordan)
• be409ae: Optimise raw MWA data reading. (Christopher H. Jordan)
• 5698b62: Beam code overhaul. (Christopher H. Jordan)
• 4006294: Allow Aman's dipole amps to be ingested. (Christopher H. Jordan)
• f4718dc: HIP support. (Christopher H. Jordan)
• 1e36c75: Release v0.3.0 (Christopher H. Jordan)

v0.2.1

Commits

• dd5cd20: Merge subcrates into a single library crate. (Christopher H. Jordan)
• 2ef57f6: Bug fixes for reading RTS solutions. (Christopher H. Jordan)
• 9f5eb3a: Stop capping spectral indices. (Christopher H. Jordan)
• 9f1beee: Improve MS timestep metadata. Fix solutions-apply. (Christopher H. Jordan)
• 5cc5c40: Don't allow picket fence obs to be calibrated. (Christopher H. Jordan)
• 7dca1b4: Map metafits tiles to uvfits/MS tiles. (Christopher H. Jordan)
• da883ce: More improvements to MS-reading code. (Christopher H. Jordan)
• 3d54ffb: Update online documentation. (Christopher H. Jordan)
• bf297fd: Tidy up the formatting of dipole-gains. (Christopher H. Jordan)
• cc66e1c: Improve default raw MWA data fine channel flags. (Christopher H. Jordan)
• 8e1ba5f: Improve logging. (Christopher H. Jordan)
• db6a56d: Add a test for multiple-timeblock calibration. (Christopher H. Jordan)
• 0e5126a: v0.2.1 (Christopher H. Jordan)

v0.2.0

Commits

• 6a51036: Enable curved power laws in GPU modelling. (Christopher H. Jordan)
• a23e4e5: Report flux-density type counts in srclist-verify. (Christopher H. Jordan)
• a6432aa: Fix srclist compatibility. (Christopher H. Jordan)
• eafc16b: Reject source list NaNs. (Christopher H. Jordan)
• 99deb4a: Fix a plotting bug. (Christopher H. Jordan)
• 8fd1220: move UvfitsWriter into Marlu (Dev McElhinney) #10
• 5b6447c: simulated weights use weight factor (Dev McElhinney) #10
• 214741d: update lock file (Dev McElhinney) #10
• 3fec33e: tests evaluate coverage better (Dev McElhinney) #10
• da94883: explicit conversion to u16 (Dev McElhinney) #10
• 434575f: re-establish write_then_read_uvfits tests (Dev McElhinney) #10
• 54a7824: coverage tweaks for simulate-vis (Dev McElhinney) #10
• fba2ff2: improve calibrate param coverage (Dev McElhinney) #10
• e0c99d1: remove hyperdrive UvfitsWriteError (Dev McElhinney) #10
• 15e9f65: validate di-cal writes uvfits, no autos (Dev McElhinney) #10
• 075ebf0: formatting, linting (Dev McElhinney) #10
• 1903251: 🐛 fix averaging factor stacking edge case (Dev McElhinney) #10
• e32697c: ✅ implement visibility output averaging (Dev McElhinney) #10
• 20b03af: Small tweaks, mostly import statements. (Christopher H. Jordan) #10
• 6b585ed: use 3d array position in calibrate params (Dev McElhinney) #11
• 1d2c5df: use array_position: LatLngHeight in ObsContext (Dev McElhinney) #11
• b227282: use array_position:LatLngHeight in simulate_vis (Dev McElhinney) #11
• 6a5d916: cargo update (Dev McElhinney) #11
• 323ee16: preference params array_pos over obs_context (Dev McElhinney) #11
• c778841: allow negative array coordinates (Dev McElhinney) #11
• e92a26f: bump Birli,Marlu 0.6.1, remove patch (Dev McElhinney) #10
• 98c75dc: don't panic if time or frequency res not specified (Dev McElhinney) #10
• 16d004a: Fix issues surrounding multiple timeblocks. (Christopher H. Jordan) #12
• 5781f9a: Move vis weighting outside calibration. (Christopher H. Jordan) #12
• 70a623f: Remove autocorrelation considerations for DI cal. (Christopher H. Jordan) #12
• 90c2c21: Fix issues around averaging factors. (Christopher H. Jordan) #12
• 4ecc9dc: remove test file after check (Dev McElhinney) #12
• 86925b4: ✅ measurement set output for calibrate vis (Dev McElhinney) #12
• 3aee42a: formalize guessing of time_res/freq_res (Dev McElhinney) #12
• a0fb55f: measurement set writing for simulate-vis (Dev McElhinney) #12
• b61be74: refactor guess time/freq res into guess interval (Dev McElhinney) #12
• 31de75d: uvfits timestep reading issue (Dev McElhinney) #12
• 356ef3f: improve precision of uvfits reading code (Dev McElhinney) #12
• f0cf606: fix uvfits reading quantization issue (Dev McElhinney) #12
• ed7ed93: cal vis out time avg for sparse time selection (Dev McElhinney) #12
• 8fb5ed3: 📝 fix misaligned bullet points (Dev McElhinney) #12
• e873302: use correct tempdir for vis writing (Dev McElhinney) #12
• dd16697: add support for reading marlu-made measurement sets (Dev McElhinney) #12
• d5690df: Small fixes. (Christopher H. Jordan) #12
• 660f834: Tidy up a bunch of code, particularly test code. (Christopher H. Jordan)
• 7abc3f7: Allow select sources to be subtracted. (Christopher H. Jordan)
• 8dc39f8: Report calibration messages more coherently. (Christopher H. Jordan)
• 0d38079: Add solutions-apply and make many changes. (Christopher H. Jordan)
• a8158c8: Error overhaul and add a documentation mdBook. (Christopher H. Jordan)
• 05fad42: Fix an issue with --collapse-into-single-source (Christopher H. Jordan)
• 0bf83c8: Make uvfits reading more robust. (Christopher H. Jordan)
• a8d2e83: Add a use-all-timesteps flag. (Christopher H. Jordan)
• b9880b6: Display CUDA device information. (Christopher H. Jordan)
• 697364b: Fix a unity dipole gains bug. (Christopher H. Jordan)
• fd10a6b: Update dependencies. (Christopher H. Jordan)
• 4b24298: Allow multiple CUDA arches and SMs to be compiled. (Christopher H. Jordan)
• d58c520: Add a feature to open libfontconfig at runtime. (Christopher H. Jordan)
• 89d91f1: Bump version. (Christopher H. Jordan)

v0.2.0-alpha.11

Commits

• f2b9686: Add a benchmark on calibration. (Christopher H. Jordan)
• 662aaaa: Use weights properly. Test uvfits/ms data reading. (Christopher H. Jordan)
• 1db64ce: Remove unnecessary depedencies from CUDA crate. (Christopher H. Jordan)
• cf495a7: Remove list->power law conversion code. (Christopher H. Jordan)
• 2dd7eb7: Use vec1 for sky model structs. (Christopher H. Jordan)
• 8b823f0: Add benchmark on estimating 'list' flux densities. (Christopher H. Jordan)
• 1e6c54f: Overhaul "list" flux density estimation. (Christopher H. Jordan)
• be0a804: Fix into_cal_sols. (Christopher H. Jordan)
• 89d7879: Use Birli 0.5.1 for PFB correction. (Christopher H. Jordan)
• c00174e: Improve error handling of bad raw data. (Christopher H. Jordan)
• 0abc675: v0.2.0-alpha.11 (Christopher H. Jordan)

v0.2.0-alpha.10

This commit is *huge* on top of main. It implements DI calibration with
many features around it, although there are still a few things to do
before we can more onto beta releases.