SD_Toolbox

Checkout Tools
  • last updated a few seconds ago
Constraints
Constraints: committers
 
Constraints: files
Constraints: dates
Fixed gridded data and lon/lat nearest neighbor which would lead to index out of range.
Added stations key to the output data. Modified get unique indices to avoid issues when filtering lists.
Added logic to handle the nearest neighbors for non-gridded data.

Now we also have unique stations and their mapping to the final stations.

Switched era5 msl key-value.
added gridded data test reading

updated progress extraction
added testing
  1. /trunk/doc/SDToolBox - Test Plan/testAntolinez
Modified code for stratified k-folds and created custom scorers.

Related to Issue [SDTOOLBOX-41]

    • -67
    • +68
    /trunk/SDToolBox/statistical_model.py
Corrected Robutness method.

Related to Issue [SDTOOLBOX-44]

    • -10
    • +82
    /trunk/SDToolBox/statistical_model.py
Enabled code for BIAS and Variance correlation.

Related to Issue [SDTOOLBOX-14]

robutness_metrics and headers for a few other functions.

Related to Issue [SDTOOLBOX-43]

    • -2
    • +45
    /trunk/SDToolBox/statistical_model.py
Implemented methods based on the scikit documentation.

Related to Issue [SDTOOLBOX-42]

    • -0
    • +32
    /trunk/SDToolBox/statistical_model.py
Time series splits.

Related to Issue [SDTOOLBOX-41]

    • -5
    • +34
    /trunk/SDToolBox/statistical_model.py
Fixed training / validation percentage. Added test.

Related to Issue [SDTOOLBOX-41]

Simple statistical model.

Related to Issue [SDTOOLBOX-41]

    • -0
    • +48
    /trunk/SDToolBox/statistical_model.py
Fixed returning NN masked indexes.

Related to Issue [SDTOOLBOX-35]

    • -60
    • +47
    /trunk/SDToolBox/extract_data.py
Fixed failing tests.
Now it is needed to explicity specify the model variable datatype.

Related to Issue [SDTOOLBOX-36]

Implemented logic for using a mask while applying nn.

* Extended NN algorithm with an eucledian approach.

* Moved input validations as they should not be in the constructor.

Related to Issue [SDTOOLBOX-35]

    • -17
    • +129
    /trunk/SDToolBox/extract_data.py
Removed unnecessary file.
Added aggregation methods. could not set the values correctly so we are returning the predictor in a tuple.

Related to Issue [SDTOOLBOX-32]

    • -59
    • +155
    /trunk/tests/test_predictor_definition.py
    • -59
    • +136
    /trunk/SDToolBox/predictor_definition.py
Fixed review remarks

Related to Issue [SDTOOLBOX-33]

Added input coordinates as lon lat dictionary with tests

Related to Issue [SDTOOLBOX-33]

    • -57
    • +218
    /trunk/tests/test_input_data.py
Added input coordinates as lon lat dictionary of arrays to be tested

Related to Issue [SDTOOLBOX-33]

Intermediate commit.

Related to Issue [SDTOOLBOX-32]

    • -190
    • +2
    /trunk/SDToolBox/data_processing.py
    • -159
    • +13
    /trunk/tests/test_data_processing.py
    • -0
    • +195
    /trunk/tests/test_predictor_definition.py
    • -0
    • +281
    /trunk/SDToolBox/predictor_definition.py
    • -0
    • +60
    /trunk/tests/TestHelper.py
added check on years InputData parameters to accept nparray as long as list

Related to Issue [SDTOOLBOX-34]

Fixed unit and system test for grid spatial resampling.

Related to Issue [SDTOOLBOX-28]

    • -4
    • +39
    /trunk/tests/test_data_processing.py
Reworked on the grid spatial resampling, fixed failing tests for time resampling.

Related to Issue [SDTOOLBOX-28]

    • -22
    • +11
    /trunk/SDToolBox/data_processing.py
    • -7
    • +114
    /trunk/tests/test_data_processing.py
Updated environment to include conda-netcdf4 which removes the hdf5 dependency.
Update of setup and environment to avoid HDF5 error on TeamCity