slices = [slice(None)] * 2
start_index = field.reversemap(time=startdate)[field.time_dim].start
end_index = field.reversemap(time=enddate)[field.time_dim].start
slices[field.time_dim] = slice(start_index, end_index + 1)
predictand_data = field.variables[0][slices]
print "predictand_data.shape = ", predictand_data.shape
locations = predictands["sources"][predictands["sources"].keys()[0]]["fields"]["tasmax"].features()
# print locations
# for location in locations['features']:
# print location
locations = locations["features"][20:22]
pca = PCA_sklearn()
pca.fit(predictors, locations, startdate=startdate, enddate=enddate)
# pca.save('pca.nc')
# pca.load('pca.nc')
reduced_predictors = pca.transform(predictors, startdate=startdate, enddate=enddate)
vmin = 0
vmax = 50
mappable = ScalarMappable(cmap="Blues")
mappable.set_array(np.arange(vmin, vmax, 0.1))
mappable.set_clim((vmin, vmax))
id = 20
for pred in reduced_predictors:
tree = BinaryTree(pred, maxdepth=10)
predictand_data = field.variables[0][slices] print 'predictand_data.shape = ', predictand_data.shape locations = predictands['sources'][predictands['sources'].keys()[0]]['fields']['pr'].features() locations = locations['features'][20:22] if method == 'pls': pls = PLS_sklearn() pls.fit(predictors, predictand_data, locations, log=True, startdate=startdate, enddate=enddate) reduced_predictors = pls.transform(predictors, startdate=startdate, enddate=enddate) reduced_predictors_subset1 = pls.transform(predictors, startdate=startdate, enddate=enddate, months=[6,7,8]) reduced_predictors_subset2 = pls.transform(predictors, startdate=startdate, enddate=enddate, months=[12,1,2]) if method == 'pca': pca = PCA_sklearn() pca.fit(predictors, locations, startdate=startdate, enddate=enddate) reduced_predictors = pca.transform(predictors, startdate=startdate, enddate=enddate) reduced_predictors_subset1 = pca.transform(predictors, startdate=startdate, enddate=enddate, months=[6,7,8]) reduced_predictors_subset2 = pca.transform(predictors, startdate=startdate, enddate=enddate, months=[12,1,2]) id = 20 for pred in reduced_predictors: ranges = zip(np.min(pred, axis=0), np.max(pred, axis=0)) tree = BinaryTree(pred, ranges=ranges, maxdepth=10) tree_subset1 = BinaryTree(reduced_predictors_subset1[id-20], ranges=ranges, maxdepth=10) tree_subset2 = BinaryTree(reduced_predictors_subset2[id-20], ranges=ranges, maxdepth=10)
