Support Vector Regression cannot handle single parameter estimations

[Martin Danner]

In the Regression Workflow, choosing the SVR as algorithm, it is set to Sklearn’s “MultiRegressorOutput” by default. An error is returned (Error: y must have at least two dimensions for multi-putput regression but has only one) when trying to use your tutorial with EnMAP_Sonoma.bsq and Biomass_Sonoma.shp with SVR instead of RandomForestRegressor.

[Andreas Janz]

As a quick workaround, just replace the last line of code with:

‌

estimator = scaledAndTunedSVR

‌

[Andreas Janz]

resolves #480

→ <<cset 55f20faeafdc>>