Types of changes
- 'Added' for new features.
- 'Changed' for changes in existing functionality.
- 'Deprecated' for soon-to-be removed features.
- 'Removed' for now removed features.
- 'Fixed' for any bug fixes.
- 'Security' in case of vulnerabilities.
Publication on the pypi.
- v25 sub-module since V25 tools are only used by FAIROpro atm, i.e. they are better situated there
- signalsteps sub-module since unused
- move from pyenv / poetry to uv
- avgbinmap.bin_y_of_t, keyword 'return_type' renamed to 'aggregation'
- geo: distance from geopy wrapper now returns array. Total distance can easily be calculated as its sum.
- txt_2_dict: fix 'preserve_empty' to actually remove surrounding delimiters
- v25_config: add H2O config (commented out)
- misc: simplify clean_path, add PCOLOR class
- drop Python 3.9 support
- add tests:
- avgbinmap:
- 10s binning,
- bin y of t
- pd resample, np reduceat
- moving averagers
- filters:
- jumps
- geo:
- sza (comparison pysolar with own impl.)
- interpolation:
- pd dataframe interpolation
- misc:
- list change element index
- find first elem
- avgbinmap:
- remove:
- avgbinmap.moving_avg (unused, buggy)
- some 'isinstance' checks which should be covered by type annotations
- misc.checkbytes_lt128 since this can be done with str.isascii() method
- timecorr.get_tcorr_parms and .apply_tcorr_parms
- avgbinmap, binning: use namedtuple to hold time vector binning results
- filter, jumps: use namedtuple to hold filter results
- timecorr: more namedtuples
- clarify type annotations for jump filter funcs
- lof filter: fix random seed
- make pure functions, do not modify input:
- interpolation, pd and pl Series
list_change_elem_index
- rename:
list_chng_elem_index
tolist_change_elem_index
posix_2_mdns
tounixtime_2_mdns
dtstr_2_posix
todtstr_2_unixtime
timecorr.time_correction
totimecorr.correct_time
txt2dict.txt_2_dict_simple
totxt2dict.txt_2_dict
- simplify
dtstr_2_mdns
; now wrapsdtobj_2_mdns
txt2dict.txt_2_dict
:- now returns a namedtuple with filepath, file header and data
- removed option to parse elements to float. this should be done in post-processing (list(map(T, col)))
- v25:
- use
txt_2_dict
- move "internal" stuff to misc
- use
- cleanup and test pd_Series interpolation
- fix readthedocs generation
- satisfy numpy 2.0 deprecations
- misc tools add path cleaner
- add LOF filter
- time corr, pl filters: replace comparison/fill with integers with vanilla Python timedelta
- NASA Ames 1001, reader: set ensure_ascii to False by default
- NASA Ames 1001, writer: set encoding to utf-8 by default
- na1001: revise parameter names of df output functions (to Python standard)
- na1001: df output functions add option to clean column names
- v25: revise line cleaning in file-collector function
- change license to LGPLv3
- enable support for Python 3.12
- v25 cleaner: add removal of ragged lines
- expose individual modules via
__all__
- readme and workflow tweaks
- adapt code to numpy 1.26 changes (xcorr-timelag function)
- na1001 / FFI1001: use today's date as default RDATE
- function 'del_at_edge' is now 'extend_mask' and can be used to extend the 'True' elements in a boolean mask (tested)
- timeconversion fix bug: xrtime_to_mdns now returns float dtype array as intended
- na1001 fix bug: prescribe FFI to be 1001
- na1001 refactor:
- better pyright integration: define FFI1001 keys in
__init__
with defaults, to describe the types - removed
__KEYS
and _show; now part of__str__
and__repr__
where needed - keys
HEADER
andSRC
are now prefixed with and underscore to indicate private attrs - add a setter for
V
, as dependent variables are internally stored in attr_V
- better pyright integration: define FFI1001 keys in
- Python 3.11 compatible versions of numba JIT-compiled functions
- drop Python 3.8 support
- avgbinmap / mean_day_frac: bugfix corner case mean day fraction is almost zero but < 0
- avgbinmap / add scipy based version of mean angle function
- revised v25 data cleaner
- numberstring: change output format of NumStr format analyser (to be used in f-string)
- na1001: add option to create FFI1001 from buffered IO
- add option to make polars df from na1001 class instance
- add 'HEADER' key to na1001 class to contain all the lines from the file header as a list
- interpolation functions, esp. for polars Series / df
- add
polars
helper functions (timecorr: filter non-monotonic datetime series) - add
calc_shift
function, used for FAIRO cl timestamp re-gridding
- NASA Ames 1001 backend updates (file format checks etc.)
- enable installation on Python 3.10
- change of some function names in V25 tools to idiomatic ones
- change of class name pyfuppes.na1001.ffi_1001 to .FFI1001
- include V25 tools
- update documentation
- update cross-correlation time lag calculator
- add pd.Series interpolation
- re-introduce
na1001
- add sphinx docs
- exclude
na1001