def __init__(self,
df=None,
data_name="empty",
data_type="standard",
column_spec=None):
self.df = df
self.data_name = data_name
self.prior_idx = np.array([], dtype=int)
if df is None:
self.column_spec = {}
return
self.max_idx = max(df.index.values) + 1
# Infer column specifications if it is not given.
if column_spec is None:
self.column_spec = {}
for col_name in list(df):
data_type = type_from_column(col_name, COLUMN_DEFINITIONS)
if data_type is not None:
self.column_spec[data_type] = col_name
else:
self.column_spec = column_spec
if "included" not in self.column_spec:
self.column_spec["included"] = "included"
if data_type == "included":
self.labels = np.ones(len(self), dtype=int)
if data_type == "excluded":
self.labels = np.zeros(len(self), dtype=int)
if data_type == "prior":
self.prior_idx = df.index.values
def __init__(self, df=None, data_name="empty", data_type="standard",
column_spec=None):
self.df = df
self.data_name = data_name
self.prior_idx = []
if df is None:
self.column_spec = {}
return
if data_type == "included":
self.labels = np.ones(len(self))
if data_type == "excluded":
self.labels = np.zeros(len(self))
if data_type == "prior":
self.prior_idx = df.index.values
self.max_idx = max(df.index.values) + 1
if column_spec is None:
self.column_spec = {}
for col_name in list(df):
data_type = type_from_column(col_name, COLUMN_DEFINITIONS)
if data_type is not None:
self.column_spec[data_type] = col_name
else:
self.column_spec = column_spec
if "final_included" not in self.column_spec:
self.column_spec["final_included"] = "final_included"