from data_utils import cast_table
from Orange.regression.earth import EarthLearner
import attr_lists
if globals().get('in_data'):
out_data = cast_table(in_data, attr_selector=)
out_classifier = EarthLearner(out_data,
degree=1,
terms=30,
penalty=1.0,
thresh=0.001,
min_span=0,
new_var_penalty=1,
fast_k=20,
fast_beta=1,
store_instances=False)
out_classifier.name = 'rv_Earth'
import Orange
from Orange import orange
from Orange.regression.earth import EarthLearner
attrs = [attribute.name for attribute in in_data.domain.attributes if attribute.name.startswith('d_') and str(attribute.var_type) == 'Continuous']
rel_attrs = []
for attr in attrs:
attr_name = 'w_per_' + attr
rel_attrs.append(orange.FloatVariable(attr_name, getValueFrom=lambda i, r, n=attr: (i[n] and i[n] > 0.0 and i['d_word_count'] / i[n]) or 0.0))
new_domain = Orange.data.Domain(rel_attrs, in_data.domain.class_var)
new_domain.addmetas(in_data.domain.getmetas())
out_data = orange.ExampleTable(new_domain, in_data)
out_classifier = EarthLearner(out_data,
degree=1,
terms=30,
penalty=1.0,
thresh=0.001,
min_span=0,
new_var_penalty=1,
fast_k=20,
fast_beta=1,
store_instances=False)
out_classifier.name = 'rel_d_Earth'
