Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Bring branch up to date with master (#579)
* Temporarily specifying pandas 2.0.0 dependency for testing purposes * Disabling akebono tests....works in my environment, looks like connection/download problems in github test environment are still an issue * generate gse2sse file and skeleton; update jwu_testto call gse2sse * Removing pandas 2.0.0 dependency (it seems to work fine) Re-enabling akebono tests to verify that the test failure there is not related to pandas 2.0.0. * Disabling akebono tests again -- test failure is unrelated to pandas 2.0.0, perhaps triggering DDOS protections with too many connections in a short time period? * Apply cotrans to sun/moon positions if necessary Apply earth-moon translation when working with positions Transpose rotation matrices if doing inverse transform (pyspedas tvector_rotate does not support invert keyword) * Switch to using CDF for IDL-Python validation (still not picking up coordinate system metadata, though) Set coordinate system explicitly where needed Add GSE to SSE rotation matrix to test variables * bumping required version of pytplot to get EIS bug fix also bumping version to get this into PyPI * fixing bug with repeated calls to FEEPS PADs with different time ranges * Adding slice_x and slice_norm keywords for MMS * Adding more 2D slices tests for MMS * Adding calls to plot the EIS data after it's loaded to the test suite (to catch any metadata regressions inside tplot) * Adding calls to plot the FEEPS data after it's loaded to the test suite (to catch any metadata regressions inside tplot) * bumping version again * Replaced first batch of tplot math/utility routines with wrappers that forward to pytplot versions. * Slight rewrite of test case * Bump required version of pytplot, to pick up relocated utilities * Temporarily (we hope!) commenting out HAPI tests, due to server problems at SPDF. * Restoring hapi tests, SPDF server is back online * Add WAVES L3 routines, fix impact STB * Downgraded the "Coordinate system not found" message from an error to a debug message. * Explicitly set coordinate system metadata in the SLP load procedure (at least until the data or master CDFs are updated) * Set the coordinate system in the output tplot variable. * Fixed a few typos; no longer need to set SLP coordinate systems explicitly; add tests for coordinate system metadata in output variables. * Add README example for WAVES * Pass sample times to np.gradient() to calculate a time derivative. * Add variable_type string parameter (default "Other"), if "pos", use lunar position as an offset value; if "vel", use lunar velocity; anything else, equivalent to rotate_only=True. * Add variable_type parameter to flag position and velocity transforms; relax velocity comparison tolerance to 1 m/sec for validation. * Trying to get coverage stats for pytplot using source_pkgs option. Also, can probably leave /home/travis/ out of the config. * improving documentation * Ignore QtPlotter directory for coverage stats * Adding geopack tests to github action * Add option to change fill_value in scipy.interpolate.interp1d * Renamed rotate_only to rotation_only to match IDL version. * Added tests for rotation-only transforms, and tests using FGM spin fit data instead of state data. * Migrated another group of tplot tools to pytplot package, replaced with wrappers that forward to the pytplot versions. * Version bump on pytplot requirement to pick up migrated tplot tools * Version bump to pick up missing imports from pytplot * Added forwarding wrappers for time_float_one and time_string_one * add sse2sel.py; fail test in tests_lunar_cotrans.py; * Removed position and velocity offset calculations, added a tplot variable for sel_y_sse for ease of testing and debugging. * Removed variable_type and rotation_only parameters (always a rotate-only transform) Added more sse<->sel tests * First batch of updated imports, for tools that moved from pyspedas.utilities to pytplot. * Next batch of import updates, for tools that have been moved to pytplot.tplot_math * Changed some np.sum calls to nansum (but not all -- some calls resulted in mismatched IDL and Python results). Uses bottleneck version of nansum if available. * Added relative tolerances to test comparisons, to catch potential differences in small output values. * Changed tplot_names() calls to data_exists() to reduce noisy output. * Changed print calls to logging calls; separate tests for bad input and output coordinate systems. * Check validity of input and output coordinate systems separately, more informative diagnostic messages. * Changed print calls to logging calls. * Removed logging messages for importing bottleneck vs. numpy version of nansum; output can be confusing, since twavpol is imported into pyspedas by default. * Better organization of imports * Simplifying matrix squaring a bit more * Added themis lunar cotrans tests * Added pytplot forwarding wrappers to coverage exclusion list....there are no remaining references to them in pyspedas proper. * Added Spinmodel object to import list * Added tvector_rotate to cotrans imports * Added tests for cotrans_set_coord not clobbering 'units' in data_att * Do not clobber other entries in data_att structure * Cleaned up imports * Cleaned up imports; ensure that input metadata is carried through intermediate variables * Cleaned up imports * Cleaned up imports * Cleaned up imports; switched back to tplot save file for baseline data; added tests to ensure 'units' metadata not getting clobbered in cotrans operations * Bump pytplot-mpl-temp required version to 2.1.32 * Bumping required pytplot version to 2.1.33, to pick up correct matplotlib requirement * Version bump for pytplot to resolve matplotlib version conflict * Version bump for pypi update * Try running the tests under Python 3.8....we have at least one backward compatibility issue, need to see if there are others. * Use backward compatible type annotation for spinmodel dictionary. * now using time_clip from pytplot * Fixing off-by-one bug * Updated to support automatic determination of the size of the Hanning window, using the same method as the IDL routine * Version bump to get Python 3.8 compatibility fix into pypi * Rolling out new version to PyPI to get dynamic power spectra bug fix * Adding docstring * updated RBSP RBSPICE routines to use bottleneck, if it's available * Fixing a crash that occurs when the user uses a decimal in the energy range of a PAD, and updating the test suite * New release, to get the latest RBSP bug fix into PyPI * Fixing a few more issues * another version bump for RBSP bug fixes * fix for issue with hourly OMNI data (#470) * version bump to get OMNI hourly bug fix into PyPI * bumping required version of PyTplot, to test new astropy units functionality * bumping required version of PyTplot, to get regression bug fix * Now setting units for HPCA angles in flux/PSD data * Performance improvements for FEEPS (get_data -> get; avoids unnecessarily converting datetime64s to floats) * Adding test for MMS LMN transformations Also adding the MMS cotrans tests to the test suite, including installation of spacepy (for the tests only) * Fixing crash when input variables don't exist * increasing code coverage for MMS cotrans functionality * Fixing crash with MMS FEEPS SITL data * Bumping version to get FEEPS bug fixes into PyPI * more FEEPS SITL bug fixes * another version bump to get FEEPS SITL bug fixes into PyPI * Now assuming datetime objects in the trange are in UTC (should fix #487) Also adding a return_times keyword to return the times * bumping version to get dailynames datetime fix into PyPI * fixing issues with EIS load routine * moving EIS fix to FEEPS as well * Update minvar_matrix_make.py Removed the unnecessary 'v' variable from the minvar_matrix function. This variable was not needed and was causing issues when `store_data`. * Update requirements.txt cdflib 1.0.0 introduces breaking changes * message for pyqt5 pyqt5 is not longer required, so we need to ask the user to install it, if it is not already installed. * Now using logging instead of print * Add STEREO beacon data, unit test, add assertion error for an instrument passed to stereo.load * Fixing crash with RBSP ESRHELT data * Bumping version for RBSP ESRHELT bug fix * Added last_version parameter to psp load routines * Now forcing latest major version for EDP data due to mixture of new/old files at the SDC * Adding support for Solar Orbiter L3 SWA HIS data * bumping version to get latest updates into PyPI * Tests for pytplot timespan and tlimit routines, and other issues pertaining to time representation. Probably not suited for automated testing until we can script plot output comparisons. * Version bump to 2.1.40, to pick up tlimit and timespan fixes, netcdf2tplot fix. * New version of CDAWeb GUI. New GUI version uses Tkinter, which is included in python. No additional requirements are needed. PyQt5 is no longer used or required. * minor update to mms_qcotrans tests * Minor changes to the GUI Improved alignment of buttons. * updates to the FEEPS test suite * adding a few more checks to increase MMS code coverage * adding error for missing trange * more MMS test suite updates * Add tests for tlimit 'last' and 'full' functions (as arguments, and as flags) * Replaced cotrans_get/set_coord with wrappers that call pytplot.get/set_coords * Replaced calls to cotrans_get/set_coord with pytplot.get/set_coords * Added new tests for pytplot get_coords, set_coords,get_units,set_units, and the wrapper functions for set_coords and get_coords * Version bump to pytplot-mpl-temp 2.1.41, to pick up attribute getters/setters * Disabling the HAPI tests, due to the CDAWeb HAPI server being offline temporarily * Autoload support data (spin axis) rather than passing support data parameters. Added probe and * Autoload support data (spin axis) rather than passing support data parameters. Added probe and use_spinaxis_corrections parameters to load and use the right support variables. * Autoload support data (spin model), added probe and eclipse_correction_level parameters to load and use the right spin model. * Autoload support data (sun and moon positions, SEL coordinate axes), rather than passing as parameters. * Updated tests, now that ssl2dsl, dsl2gse, gse2sse, and sse2sel now autoload the support data they need. * Removed support data parameters no longer needed * Removed support data parameters no longer needed * Add test of various timebar string formats * Version bump to 2.1.42 to pick up timebar changes * Re-enabling HAPI tests, now that SPDF's HAPI service is back online * Fixing crash due to incompatible files at the SDC (mixture of v6 and v7 files) * Fixing crash when PA data can't be found * Adding flag to assume no wild cards in the URL/filename * minor update, fixing typo in the docstring * Adding initial code for grabbing and caching the latest burst intervals from the SDC * Adding test for routine that updates the SDC burst segment intervals * moving cdflib version requirement over from requirements.txt * bumping version to get new cdflib requirement into PyPI * When the downloaded file cannot be opened, try again. Added downloading the file for a second time, if it is a CDF or netCDF file and cannot be opened correctly. * improving the error handling in mms_qcotrans (fixes #515) * Fixing problem with latest FEEPS masking table for MMS1 (extra column appended to the end); fixes #545 * bumping version to get FEEPS bug fix into PyPI * Fix sqtn_rfs_V1V2 non-public download * fixing bug time clipping spin averaged FEEPS spectra (should fix #498) * adding regression test for FEEPS time clipping bug * Now deleting the temporary file when NamedTemporaryFile is closed * Version bump to get temp file bug fix into PyPI * Now loading burst intervals from the SDC by default; should fix #378 * Updating burst mode data bar test * Disabling checks for Solar Orbiter RPW FLUX_DENSITY# variables (these variables no longer exist in the v05 CDFs * Slightly different fix for #555, to fix #562 (now deleting the temporary file using os.unlink instead of relying on NamedTemporaryFile) * Slightly different fix for #555, to fix #562 (now deleting the temporary file using os.unlink instead of relying on NamedTemporaryFile) * Now preserving metadata after interpolation; should fix #475 * version bump * fixing regression due to units now being returned by np.arccos * version bump to get FEEPS gyrophase bug fix into PyPI * Temporarily increase logging verbosity to detect any new deprecation warnings * Temporarily removing basemap installation and secs test, to allow testing with numpy 1.25 * Backing out inadvertent commit * Temporarily requiring numpy 1.25.0 to find new deprecations in test suites * Use Python 3.9 to allow use of numpy 1.25 * Temporarily commenting out spacepy installation and mms cotrans tests to run under python 3.9/numpy 1.25 * Fixing deprecated singleton array -> scalar implicit conversion reported by numpy 1.25 * Fixing deprecated singleton array -> scalar implicit conversion reported by numpy 1.25, cleaning up indexing a bit by flattening index arrays * Fixing deprecated singleton array -> scalar implicit conversion reported by numpy 1.25. * Revert to numpy >= 1.19.5, bump pytplot version to 2.1.46 to pick up UNIT_PTR fix * Revert to Python 3.8 environment w/ spacepy and basemap installations, less verbose logging * Bumping to pytplot 2.1.47 to pick up bug fix when units metadata specified as an array * rfs_hfr has /%Y/%m/ path format on SSL server * test * All Berkeley files stored in monthly directories * added barrel data loader * Changed default date to one where data is available * Restoring akebono tests; adding barrel tests * Adding BARREL_DATA_DIR to environment * Taking akebono test suite back out; connection attempts from Github timing out for some reason. * Bump required version of Python to 3.8 (due to numpy dependency). --------- Co-authored-by: Jim Lewis <[email protected]> Co-authored-by: jwu <[email protected]> Co-authored-by: supervised <[email protected]> Co-authored-by: Samuel Badman <[email protected]> Co-authored-by: Jim Lewis <[email protected]> Co-authored-by: Tien Vo <[email protected]> Co-authored-by: Beforerr <[email protected]> Co-authored-by: Nick H <[email protected]> Co-authored-by: Marc Pulupa <[email protected]> Co-authored-by: Warren Rexroad <[email protected]>
- Loading branch information
11 people
authored
Aug 25, 2023
1 parent
0dd92cf
commit 004e59d
