Added parcel mode outputs, improved docs, improved samir parame tests, minor corrections and changes.

Further improvements in optimization. Manage to create real empty output dataset and append values along the time dimension, greatly improves total run speed. Reworked the custom_input notebook for a better handling of the land cover class conversion. Added tests on parameters values to make sure they are in correct range, parameter ranges are defined in a csv file (params_samir_min_max.csv). Other small modifications.

Some optimization improvements. Potential memory leaks to look at. Many improvements in the preprocessing scripts, almost finished the parcel mode. New scripts in the prostprocessing directory. Changes in the samir parameter handling.

Improved output plotting and added a script to automatically generate a daily GIF for a chosen variable.

Added main_run_samir.py to run the samir model. Modified config_file to manage input paths. Various bug corrections in dataset dimensions, speed improvements with parallel processing parameters. Added a notebook in postprocessing to plot pixel model outputs.

Various modifications in preprocessing codes. Improved datachunking, corrected bugs in weather dataset download, small improvements in script speed. Added a notebook to help prepare landcover and soil data for pixel mode

Finished ndvi extraction scripts. Added parcel mode to weather scripts. Wrote script to convert dataframes into xarray datasets for parcel mode.

Modified input calculations. Reprojection of weather dataset needs to be handled differently for bigger images (with OTB for example). Modified the parameter class. Begining to work on samir time loop, inputs and parameters are done. Added env file. New method for calculating ET0 with xarray might have been found.

Calculation of ndvi cube and weather cube operational. Starting the samir params handling and time loop architecture