Update to version 0.7.0 #273

OnnoEbbens · 2023-10-18T14:40:42Z

No description provided.

- use chunks and dask delayed to keep memory usage efficient - get_geotop now gets raw data - add deprecationwarning to get_geotop_raw_within_extent()

- now takes geotop_ds as input to convert to layers (not extent)

- do not use sortby but check diff of y dimension and flip if necessary

- use .values instead of .data to load data into memory if using dask - undo onno changes for now w missing strat code

- add kwarg delayed, if False, load data into memory, else return data array with delayed dask arrays - add kwarg chunked, if True, chunk data array with chunks="auto" - add x,y data to vertex grid data array

- add function name in all cases - add line number when level is DEBUG

- use .values to deal with dask arrays - use get_geotop - update docstring

…top_to_ml_layers

Add

* fix for #199 * fix * return da in some situations --------- Co-authored-by: Ruben Caljé <r.calje@artesia-water.nl>

Small fix for Error: Assigning CRS to a GeoDataFrame without a geometry column is not supported

nlmod/dims/layers.py

nlmod/gwf/gwf.py

dbrakenhoff · 2023-10-20T09:27:18Z

nlmod/gwf/recharge.py

+        when it is None. The default is None.
+    finf :float, str or array-like
+        The applied infiltration rate of the UZF cell (:math:`LT^{-1}`). When passed as
+        string, the array is obtained from ds. The default is "recharge".


Do we want to name this "recharge" by default or should it be precipitation? Recharge implies we already include evaporation somehow?

We can also set the default to None so the user has to decide what data should be used?

nlmod/read/rws.py

nlmod/util.py

dbrakenhoff

Lot's of great improvements :)! I made a few minor comments, nothing major. These could also be converted into issues to deal with in a next version.

dbrakenhoff · 2023-10-20T10:22:19Z

I think we should just use our admin powers to ignore issues on codacy, or somehow add some of the issues it raises to the ignore list. Most of them aren't really an issue or worth fixing.

dbrakenhoff · 2023-10-21T13:02:33Z

Has #261 been fixed yet? It would be good to get that one done before this release I think.

OnnoEbbens · 2023-10-23T08:24:47Z

I fixed most of the codacy errors. Some ideas/remarks/questions:

I disabled the too_complex pattern and the too_many_branches patterns because that are often hard to solve. However I think the checks are good indicators of code that is hard to read so I feel conflicted about disabling them.
While looking through some files I notice many functions miss a docstring. See functions in dcs.py, geotop.py and regis.py. I think we should enforce docstrings.
@bdestombe: {} is preferred over dict() according to the codacy style checker. I already fixed this but I think it is good to know for future commits.
@dbrakenhoff: can you check this codacy error, maybe replace pass with ...:

nlmod/util.py
Medium icon Medium Code Style
Unnecessary pass statement

pass

@dbrakenhoff: can you check this codacy error, I cannot figure out if we can ignore this:

nlmod/dims/base.py
Minor icon Minor Code Style
bad operand type for unary ~: NoneType (invalid-unary-operand-type)

data[lay, mask] = data[lay, ~mask][i]

@dbrakenhoff: probably the same as the previous one:

nlmod/dims/base.py
Minor icon Minor Code Style
bad operand type for unary ~: NoneType (invalid-unary-operand-type)

data[mask] = data[~mask, i]

@rubencalje: can you check this codacy error (and add a docstring for the dataframe_to_flopy_timeseries function?):

nlmod/dims/time.py
Minor icon Minor Code Style
Either all return statements in a function should return an expression, or none of them should. (inconsistent-return-statements)

def dataframe_to_flopy_timeseries(

@rubencalje: can you check this codacy error? Should we add skimage as a dependency?

nlmod/read/rws.py
Minor icon Minor Code Style
Import outside toplevel (skimage.morphology.reconstruction) (import-outside-toplevel)

from skimage.morphology import reconstruction

OnnoEbbens · 2023-10-23T08:38:07Z

Suddenly a new check: too-many_statements appeared. I decided to ignore that one

dbrakenhoff · 2023-10-23T08:38:35Z

The ~ issues can be ignored. I'm not sure how to avoid the check being triggered for these cases.

* Add nhi.py and test * Improve nhi.py * Small changes (unrelated to buisdrainage) * Add some options to modpath * Fix Issue #261 * Fix codacy stuff and improve some testst * Fix the unit of the conductance of buisdrainage * Fix test * some minor fixes - added one docstring, fixed some typos, sort imports * Add docstrings to nhi * Replace seasons by winter_name and summer_name in surface_water.py * Remove warnings and and output from unsaturated notebooks --------- Co-authored-by: Davíd Brakenhoff <d.brakenhoff@artesia-water.nl>

dbrakenhoff and others added 30 commits June 16, 2023 11:13

improve get_geotop

2c918b4

- use chunks and dask delayed to keep memory usage efficient - get_geotop now gets raw data - add deprecationwarning to get_geotop_raw_within_extent()

old get_geotop is renamed geotop_to_layermodel:

456b983

- now takes geotop_ds as input to convert to layers (not extent)

clean up old code

84593ce

improve get_regis:

62a9861

- do not use sortby but check diff of y dimension and flip if necessary

add cache support for chunked datasets

56a264e

add log msg to split_layers

68a4c1f

remove nan layer (value -127) from geo_eenheden

4ed3976

use values instead of data to deal with potential dask array

25f4f72

improve geotop:

f9cc9d7

- use .values instead of .data to load data into memory if using dask - undo onno changes for now w missing strat code

if caching with dask, close and reopen ds to ensure ds points to file

34dfdcf

Use dask delayed to read model output

5376c19

- add kwarg delayed, if False, load data into memory, else return data array with delayed dask arrays - add kwarg chunked, if True, chunk data array with chunks="auto" - add x,y data to vertex grid data array

add delayed, chunked kwargs to output methods

2179557

fix DeprecationWarning

de54e3d

improve log messages grid.py

5618671

replace modflow with mf6 in logging

758f87d

replace modflow with mf6 in logging

a70974a

more info in log message

5044c04

- add function name in all cases - add line number when level is DEBUG

docstrings

c6c1aa3

fix tests

87e7afe

fix geotop_lithok_in_cross_section

a6a26c5

- use .values to deal with dask arrays - use get_geotop - update docstring

Rename geotop_to_layermodel to to_model_layers and remove convert_geo…

471e724

…top_to_ml_layers

Some minor changes

26db2e8

Fix combination of regis and geotop

a551f55

version bump

b4b5256

Improve geotop.to_model_layers and add method insert_layer

fd588c7

Add extra legends for layer models

776a654

Add

Add Formatie van Nieuwkoop, Laagpakket van Nij Beets in geotop

ecbe766

shorten name model_dataset_created_on -> created_on

4fba381

199 nlmodgwfic slow for uniform starting head (#200)

62eadc9

* fix for #199 * fix * return da in some situations --------- Co-authored-by: Ruben Caljé <r.calje@artesia-water.nl>

Update webservices.py

961f321

Small fix for Error: Assigning CRS to a GeoDataFrame without a geometry column is not supported