Prepare release 2.8.0

[veenstrajelmer]

TODO:

• Remove input/return arguments from init_RWS and exit_RWS #106
• create callable hatyan module with __main__.py #113
• phase out analysis_peryear #115
• support glob in readts_dia #119
• expose ts.index.freq in case of multifile dia #117
• fix failing freq testcase on github #121
• prevent user interaction with get_components_from_ts but analysis instead #124
• rename comp_allyears to comp_allperiods in plot_components() #126
• extend/correct 19Ycomp4Ydia #128
• add write+read dia testcase to check if diff is within 1cm accuracy margin #137
• Pass metadata between objects #93, simplifies example scripts by not having to pass metadata manually
• add L9PFM and Q1 to xfac0 list in all example files
• Replace times_ext list and timestep_min with slice(tstart,tstop,tstep_min) #104
• from setup.cfg to pyproject.toml #151
• fix testbank for py39 and higher versions #154
• Define minimal dependency versions and add github testcase #150
• update hmc env #146
• Simplify datetimeindex for 1018 #132
• fix sonarcloud quality gate #159
• Fix deprecation warnings in pytest #153
• python -m hatyan -h does not work anymore?
• fix hanging acceptance tests #163
• plot_astrog_diff more robust #165
• increase test coverage of timeseries.py and components.py #168
• remove HWtimediff mode in calc_HWLWnumbering #170
• prevent duplicate A0 when reading hatyan 2.7.0 component file #172
• Get nodalfactors/xfac settings from component file #144 (discuss issues with JR/AB)
• Acceptance tests failing #176
• deprecate requirements.txt #179
• Update build method #180
• add code3 comment in component file #183
• move diawia acceptancetest to systemtest #185
• uncomment py312 classifier in toml file
• fix quality gate #189
• acceptance test astrog_test.py fails #166
• remove .gitattributes?
• simplify testbank with matrix #194
• Drop support for components file without metadata #173
• Update metadata with prediction #198
• Expand and generalize ddlpy_to_hatyan() #212
• add ddlpy dependency and enable testcases #207
• during release: pypi via api token instead of password dfm_tools#732
• Improve metadata in functions and component file #139 (discuss with JR/AB)
• use pandas.DataFrame.attrs instead of separate properties #217
• update readme: remove ddl part an kwk part
• Fix acceptance tests #214
• merge write_tsdia_HWLW into write_tsdia #215
• Fix FutureWarnings from pytest python 3.12 #220
• Increase test coverage #227
• From main to click #226
• increase ddlpy_helpers test coverage #230
• define __all__ in all code scripts #184
• drop assert for fu_alltimes comp-settings in prediction #233
• Re-enable proper warnings in _read_components_analysis_settings #177
• Use logging instead of prints #98
• check timezone of ddlpy with validation dataset #237
• restore script header in https://github.com/Deltares/hatyan/blob/main/tests/examples/predictie_2019_19Ycomp4Ydia_VLISSGN.py
• From makedocs to sphinx #241
• fix exposure of functions #243
• update ddl example script #247
• fix pytest-py38-mindeps workflow #249
• remove docs/hatyan folder, this is not used anymore.
• Consider moving predictie_2022_frommergedcomp_LAT.py #221
• improve support for custom xfactors #252
• Add support for diafile with multiple blocks in get_diaxycoords #254
• Add timezone info to hatyan dataframes #238
• Simplify write_tsnetcdf by deriving metadata from input dataframes #216
• check timeout acceptance: numbering_extremes and spatial_plots (after adding dfmt.plot_coastlines)
• update testbank, jupyter notebook, example scripts
• deprecate return_listoptions argument in hatyan.hatyan_core.get_const_list_hatyan() #261
• rename read/write ts functions to something consistent (remove ts) #264
• update example notebook #266
• Cleanup analysis/prediction settings #263
• decide on settings arguments method #269
• fix FutureWarning about parse_dates (pytest python 3.12) #271
• add xfactor assertion in example scripts #274
• Consider adding Components and Timeseries classes #130
• move logging of analysis settings to hatyan.analysis(), not log for each period #276
• single source of truth for analysis settings defaults #278
• deprecate components_timeshift() #280
• simplify ddlpy to hatyan conversion #283
• increase test coverage #282
• simplify timestep_min behaviour #286
• add prints of comp and ts to example notebook (in separate blocks)
• rename analysis arguments to lowercase #291
• fix sonarcloud issues by replacing numpy where with nonzero #294
• component splitting fails if CS_comps_derive components are in const_list #289
• analysis not possible for 1018 timeseries #287
• shorter variable names in example scripts to make it less messy
• cleanup plot_HWLW_validatestats() #297
• simplify public functions by deprecating some arguments #299
• Fix FutureWarnings about deprecated Y freq (pytest python 3.12) #225
• Incorrect HW in timeseries with gap #97
• run entire testbank and check acceptance test output
• update whatsnew

Overview of user impact of finished todo's:

• https://github.com/Deltares/hatyan/blob/main/docs/whats-new.md
• 113 create callable hatyan module #114
• 115 phase out analysis peryear #116
• renamed comp_allyears to comp_allperiods #127
• 93 pass metadata between objects #131
• 104 replace times ext list and timestep min #143
• 144 get nodalfactorsxfac settings from component file #175 (including support for old hatyan1 files without analysis settings)
• avoided dummy metadata upon reading non-meta component file #196 (this change makes it impossible to read haytan2<=2.7.0 component files since they do not contain STAT/PERD metadata lines, an error is raised)
• 139 improve metadata in functions and component file #218 (prevent prediction with different settings than in components metadata and prevent non-standard nodalfactors/fu_alltimes/source in component file)
• 215 merge write tsdia hwlw into write tsdia #223 (deprecated hatyan.write_tsdia_HWLW())
• 226 from main to click #229 (makes hatyan commmand available via command line directly, instead of via python -m hatyan)
• 98 improve loggingstatistics #236 (prevented most of the prints, now to be enabled with logging module)
• 241 from makedocs to sphinx #242 (improved documentation on github)
• 216 simplify write tsnetcdf by deriving metadata from input dataframes #260 (removed station/vertref/tzone_hr arguments for hatyan.write_tsnetcdf(), comes from timeseries dataframes now, tzone was removed in 238 add timezone info to hatyan dataframes #258)
• dropped ts from all read/write functions for timeseries #265 (dropped "ts" from all read/write timeseries function names)
• 263 cleanup analysisprediction settings #268 (hatyan.prediction() does not accepts settings anymore, they are derived from the components dataframe now, this prevents accidental mistakes)
• 286 simplify timestep min behaviour #288 (prediction timestep now only accepts frequency string like "10min", not 10 anymore)
• renamed CS_comps and xTx_condition_max #292 (renamed CS_comps argument for hatyan.prediction() to lowercase)
• 299 simplify public functions by deprecating some arguments #300 (deprecated comp_sec_list argument for hatyan.merge_componentgroups(), now use .iloc[] on comp_sec instead`)
• overview by comparing the example script: old (2.7.0) vs. new/simplified (main)
• improved module documentation at deltares.github.io

FEWS HWLW numbering impact:

• removed station/vertref/tzone_hr arguments for hatyan.write_tsnetcdf(): 238 add timezone info to hatyan dataframes #258 and
  216 simplify write tsnetcdf by deriving metadata from input dataframes #260
• renamed hatyan.writets_netcdf() to hatyan.write_netcdf() dropped ts from all read/write functions for timeseries #265

After release:

• build method was updated, check if data was packaged in wheel (pip install in clean env)

Follow-up: #310