def __init__(self, source, rowvaluesatt, colvaluesatt, values,
aggregator=None, nonevalue=0, sortrows=False):
"""Arguments:
- source: the data source to pull data from
- rowvaluesatt: the name of the attribute that holds the values that
appear as rows in the result
- colvaluesatt: the name of the attribute that holds the values that
appear as columns in the result
- values: the name of the attribute that holds the values to aggregate
- aggregator: the aggregator to use (see pygrametl.aggregators). If not
given, pygrametl.aggregators.Sum is used to sum the values
- nonevalue: the value to return when there is no data to aggregate.
Default: 0
- sortrows: A boolean deciding if the rows should be sorted.
Default: False
"""
self.__source = source
self.__rowvaluesatt = rowvaluesatt
self.__colvaluesatt = colvaluesatt
self.__values = values
if aggregator is None:
from pygrametl.aggregators import Sum
self.__aggregator = Sum()
else:
self.__aggregator = aggregator
self.__nonevalue = nonevalue
self.__sortrows = sortrows
self.__allcolumns = set()
self.__allrows = set()
class CrossTabbingSource(object):
"A source that produces a crosstab from another source"
def __init__(self,
source,
rowvaluesatt,
colvaluesatt,
values,
aggregator=None,
nonevalue=0,
sortrows=False):
"""Arguments:
- source: the data source to pull data from
- rowvaluesatt: the name of the attribute that holds the values that
appear as rows in the result
- colvaluesatt: the name of the attribute that holds the values that
appear as columns in the result
- values: the name of the attribute that holds the values to aggregate
- aggregator: the aggregator to use (see pygrametl.aggregators). If not
given, pygrametl.aggregators.Sum is used to sum the values
- nonevalue: the value to return when there is no data to aggregate.
Default: 0
- sortrows: A boolean deciding if the rows should be sorted.
Default: False
"""
self.__source = source
self.__rowvaluesatt = rowvaluesatt
self.__colvaluesatt = colvaluesatt
self.__values = values
if aggregator is None:
from pygrametl.aggregators import Sum
self.__aggregator = Sum()
else:
self.__aggregator = aggregator
self.__nonevalue = nonevalue
self.__sortrows = sortrows
self.__allcolumns = set()
self.__allrows = set()
def __iter__(self):
for data in self.__source: # first we iterate over all source data ...
row = data[self.__rowvaluesatt]
col = data[self.__colvaluesatt]
self.__allrows.add(row)
self.__allcolumns.add(col)
self.__aggregator.process((row, col), data[self.__values])
# ... and then we build result rows
for row in (self.__sortrows and sorted(self.__allrows) \
or self.__allrows):
res = {self.__rowvaluesatt: row}
for col in self.__allcolumns:
res[col] = \
self.__aggregator.finish((row, col), self.__nonevalue)
yield res
class CrossTabbingSource(object):
"""A source that produces a crosstab from another source"""
def __init__(self, source, rowvaluesatt, colvaluesatt, values,
aggregator=None, nonevalue=0, sortrows=False):
"""Arguments:
- source: the data source to pull data from
- rowvaluesatt: the name of the attribute that holds the values that
appear as rows in the result
- colvaluesatt: the name of the attribute that holds the values that
appear as columns in the result
- values: the name of the attribute that holds the values to aggregate
- aggregator: the aggregator to use (see pygrametl.aggregators). If not
given, pygrametl.aggregators.Sum is used to sum the values
- nonevalue: the value to return when there is no data to aggregate.
Default: 0
- sortrows: A boolean deciding if the rows should be sorted.
Default: False
"""
self.__source = source
self.__rowvaluesatt = rowvaluesatt
self.__colvaluesatt = colvaluesatt
self.__values = values
if aggregator is None:
from pygrametl.aggregators import Sum
self.__aggregator = Sum()
else:
self.__aggregator = aggregator
self.__nonevalue = nonevalue
self.__sortrows = sortrows
self.__allcolumns = set()
self.__allrows = set()
def __iter__(self):
for data in self.__source: # first we iterate over all source data ...
row = data[self.__rowvaluesatt]
col = data[self.__colvaluesatt]
self.__allrows.add(row)
self.__allcolumns.add(col)
self.__aggregator.process((row, col), data[self.__values])
# ... and then we build result rows
for row in (self.__sortrows and sorted(self.__allrows)
or self.__allrows):
res = {self.__rowvaluesatt: row}
for col in self.__allcolumns:
res[col] = \
self.__aggregator.finish((row, col), self.__nonevalue)
yield res