import verde as vd
# We'll test this on the air temperature data from Texas
data = vd.datasets.fetch_texas_wind()
coordinates = (data.longitude.values, data.latitude.values)
region = vd.get_region(coordinates)
# Use a Mercator projection for our Cartesian gridder
projection = pyproj.Proj(proj="merc", lat_ts=data.latitude.mean())
# The output grid spacing will 15 arc-minutes
spacing = 15 / 60
# This spline will automatically perform cross-validation and search for the optimal
# parameter configuration.
spline = vd.SplineCV(dampings=(1e-5, 1e-3, 1e-1), mindists=(10e3, 50e3, 100e3))
# Fit the model on the data. Under the hood, the class will perform K-fold
# cross-validation for each the 3*3=9 parameter combinations and pick the one with the
# highest R² score.
spline.fit(projection(*coordinates), data.air_temperature_c)
# We can show the best R² score obtained in the cross-validation
print("\nScore: {:.3f}".format(spline.scores_.max()))
# And then the best spline parameters that produced this high score.
print("\nBest spline configuration:")
print(" mindist:", spline.mindist_)
print(" damping:", spline.damping_)
# Now we can create a geographic grid of air temperature by providing a projection
# This type of tuning is important and should always be performed when using a new
# gridder or a new dataset. However, the above implementation requires a lot of
# coding. Fortunately, Verde provides convenience classes that perform the
# cross-validation and tuning automatically when fitting a dataset.
########################################################################################
# Cross-validated gridders
# ------------------------
#
# The :class:`verde.SplineCV` class provides a cross-validated version of
# :class:`verde.Spline`. It has almost the same interface but does all of the above
# automatically when fitting a dataset. The only difference is that you must provide a
# list of ``damping`` and ``mindist`` parameters to try instead of only a single value:
spline = vd.SplineCV(
dampings=dampings,
mindists=mindists,
)
########################################################################################
# Calling :meth:`~verde.SplineCV.fit` will run a grid search over all parameter
# combinations to find the one that maximizes the cross-validation score.
spline.fit(proj_coords, data.air_temperature_c)
########################################################################################
# The estimated best damping and mindist, as well as the cross-validation
# scores, are stored in class attributes:
print("Highest score:", spline.scores_.max())
print("Best damping:", spline.damping_)
print("Best mindist:", spline.mindist_)
# This type of tuning is important and should always be performed when using a new
# gridder or a new dataset. However, the above implementation requires a lot of
# coding. Fortunately, Verde provides convenience classes that perform the
# cross-validation and tuning automatically when fitting a dataset.
########################################################################################
# Cross-validated gridders
# ------------------------
#
# The :class:`verde.SplineCV` class provides a cross-validated version of
# :class:`verde.Spline`. It has almost the same interface but does all of the above
# automatically when fitting a dataset. The only difference is that you must provide a
# list of ``damping`` and ``mindist`` parameters to try instead of only a single value:
spline = vd.SplineCV(
dampings=[None, 1e-5, 1e-4, 1e-3, 1e-2, 1e-1],
mindists=[5e3, 10e3, 25e3, 50e3, 75e3, 100e3],
)
spline.fit(proj_coords, data.air_temperature_c)
########################################################################################
# The estimated best damping and mindist, as well as the cross-validation scores, are
# stored in class attributes:
print("Highest score:", spline.scores_.max())
print("Best damping:", spline.damping_)
print("Best mindist:", spline.mindist_)
########################################################################################
# Finally, we can make a grid with the best configuration to see how it compares to the
# default result.