def __init__(
self,
value_s,
feature_df,
feature_to_category,
feature_name="feature",
category_name="category",
feature_summary=None,
feature_summary_fun=None,
category_summary=None,
category_summary_fun=None,
min_features_per_cat=2,
max_dist=0,
chrom_len=None,
ncpus=1,
):
# todo: logging for AssertionErrors, explain what the problem is
hp.check_rod(value_s)
hp.check_feature_df(feature_df)
assert feature_name in feature_df.columns
assert (feature_summary is not None) != (
feature_summary_fun is not None
), "Specify either feature_summary OR feature_summary_fun."
assert (category_summary is not None) != (
category_summary_fun is not None
), "Specify either category_summary OR category_summary_fun."
# if chrom_len is not None:
# self.flat_index = True
# else:
self.flat_index = False
# if self.flat_index:
# self.value_s = hp.rod_to_1d(value_s, chrom_len)
# self.feature_df = hp.feature_df_to_1d(feature_df, chrom_len)
# else:
self.value_s = value_s.copy()
self.feature_df = feature_df.copy()
self.value_name = self.value_s.name
self._bind_feature_to_category(feature_to_category, feature_name, category_name)
self.prune_feature_to_category(min_features_per_cat)
self.feature_summary = feature_summary
self.category_summary = category_summary
self.feature_summary_fun = feature_summary_fun
self.category_summary_fun = category_summary_fun
# new
self.features_per_data = hp.get_features_per_data(
self.value_s, self.feature_df, feature_name=self.feature_name, max_dist=max_dist
)
self.features_per_data.index = self.features_per_data.index.droplevel(2)
self.init_rank_table = self.initital_rank_table()
self.rank_table = self.init_rank_table
self.ncpus = ncpus
def __init__(self, value_s, feature_df, feature_to_category, feature_name='feature', category_name='category',
feature_summary=None, feature_summary_fun=None,
category_summary=None, category_summary_fun=None, min_features_per_cat=2,
max_dist=0, chrom_len=None, ncpus=1):
#todo: logging for AssertionErrors, explain what the problem is
hp.check_rod(value_s)
hp.check_feature_df(feature_df)
assert feature_name in feature_df.columns
assert (feature_summary is not None) != (feature_summary_fun is not None), \
"Specify either feature_summary OR feature_summary_fun."
assert (category_summary is not None) != (category_summary_fun is not None), \
"Specify either category_summary OR category_summary_fun."
if chrom_len is not None:
self.flat_index = True
else:
self.flat_index = False
if self.flat_index:
self.value_s = hp.rod_to_1d(value_s, chrom_len)
self.feature_df = hp.feature_df_to_1d(feature_df, chrom_len)
else:
self.value_s = value_s.copy()
self.feature_df = feature_df.copy()
self.value_name = self.value_s.name
self._bind_feature_to_category(feature_to_category, feature_name, category_name)
self.prune_feature_to_category(min_features_per_cat)
self.feature_summary = feature_summary
self.category_summary = category_summary
self.feature_summary_fun = feature_summary_fun
self.category_summary_fun = category_summary_fun
self.max_dist = max_dist
#logging.debug("I am before.")
#import pdb
#logging.debug("I am here.")
#pdb.set_trace()
#logging.debug("I am after.")
self.init_rank_table = self.initital_rank_table()
self.rank_table = self.init_rank_table
self.ncpus = ncpus
