def profile_point(self, sample, reference):
buckets = reference._buckets
data = [0] * len(buckets)
if (sample is not None) and (sample in buckets):
data[buckets.index(sample)] = 1
return DiscreteHistogram(buckets, data)
def _profile_column(self, df, column):
total = df.shape[0]
try:
missing = sum(df[column].isnull())
except KeyError: # pandas raises this if the column doesn't exist
missing = total
return (column + "_missing", DiscreteHistogram(["present", "missing"], [total - missing, missing]))
def profile_column(self, df, column):
"""Profile a data frame column, returning a name and DiscreteHistogram
Parameters
----------
df : pandas.core.frame.DataFrame
A data frame to profile.
column : str
The column in the data frame to profile.
Returns
-------
dict
A dictionary from summary data names to DiscreteHistogram instances.
"""
content = df[column].value_counts(dropna=True)
keys = list(content.keys())
values = [content[k] for k in keys]
values = [v.item() for v in values]
return (column + "_histogram", DiscreteHistogram(keys, values))
def profile_column(self, df, column):
"""Profile a data frame column, returning a name and DiscreteHistogram
Parameters
----------
df : pandas.core.frame.DataFrame
A data frame to profile.
column : str
The column in the data frame to profile.
Returns
-------
dict
A dictionary from summary data names to DiscreteHistogram instances.
"""
total = df.shape[0]
try:
missing = sum(df[column].isnull())
except KeyError: # pandas raises this if the column doesn't exist
missing = total
return (column + "_missing",
DiscreteHistogram(["present", "missing"],
[total - missing, missing]))
def profile_point(self, sample, reference):
if sample is None: # Missing
return DiscreteHistogram(reference._buckets, [0, 1])
return DiscreteHistogram(reference._buckets, [1, 0])
def _profile_column(self, df, column):
content = df[column].value_counts()
keys = list(content.keys())
values = [content[k] for k in keys]
values = [v.item() for v in values]
return (column + "_histogram", DiscreteHistogram(keys, values))