def group_by(self,
queryset,
grouping_key=None,
bins=None,
bin_size=None,
**kwargs):
"""
Return a ValuesQuerySet that has been grouped by this dimension.
The group value will be available as grouping_key in the dictionaries.
The grouping key defaults to the dimension key.
If num_bins or bin_size is not provided, an estimate will be used.
.. code-block:: python
messages = dim.group_by(messages, 'value', 100)
distribution = messages.annotate(count=Count('id'))
print distribution[0]
# { 'value': 'hello', 'count': 5 }
"""
# Type checking
queryset = find_messages(queryset)
expression = self.get_grouping_expression(queryset,
bins=bins,
bin_size=bin_size,
**kwargs)
if expression is None:
# no results
return queryset.values()
if expression == self.field_name:
# We still have to map back to the requested grouping_key
return MappedValuesQuerySet.create_from(
queryset.values(expression), {expression: grouping_key})
else:
# Sometimes this step gets complicated
queryset, internal_key = self.select_grouping_expression(
queryset, expression=expression)
# Then use values to group by the grouping key.
queryset = queryset.values(internal_key)
return MappedValuesQuerySet.create_from(queryset, {
internal_key: grouping_key,
})
def render(self, queryset, desired_primary_bins=None, desired_secondary_bins=None):
"""
Given a set of messages (already filtered as necessary),
calculate the data table.
Optionally, a number of primary and secondary bins may be given.
The result is a list of dictionaries. Each
dictionary contains a key for each dimension
and a value key for the count.
"""
if not self.secondary_dimension:
# If there is only one dimension, we should be able to fall back
# on that dimension's group_by() implementation.
queryset = self.primary_dimension.group_by(queryset,
grouping_key=self.primary_dimension.key,
bins=desired_primary_bins)
return queryset.annotate(value=models.Count('id'))
else:
# Now it gets nasty...
primary_group = self.primary_dimension.get_grouping_expression(queryset,
bins=desired_primary_bins)
secondary_group = self.secondary_dimension.get_grouping_expression(queryset,
bins=desired_secondary_bins)
if primary_group is None or secondary_group is None:
# There is no data to group
return queryset.values()
queryset, internal_primary_key = self.primary_dimension.select_grouping_expression(
queryset,
primary_group)
queryset, internal_secondary_key = self.secondary_dimension.select_grouping_expression(
queryset,
secondary_group)
# Group the data
queryset = queryset.values(internal_primary_key,
internal_secondary_key)
# Count the messages
queryset = queryset.annotate(value=models.Count('id'))
# We may need to remap some fields
mapping = {}
if internal_primary_key != self.primary_dimension.key:
mapping[internal_primary_key] = self.primary_dimension.key
if internal_secondary_key != self.secondary_dimension.key:
mapping[internal_secondary_key] = self.secondary_dimension.key
if len(mapping) > 0:
return MappedValuesQuerySet.create_from(queryset, mapping)
else:
return queryset
def group_by(self, queryset, grouping_key=None, bins=None, bin_size=None, **kwargs):
"""
Return a ValuesQuerySet that has been grouped by this dimension.
The group value will be available as grouping_key in the dictionaries.
The grouping key defaults to the dimension key.
If num_bins or bin_size is not provided, an estimate will be used.
.. code-block:: python
messages = dim.group_by(messages, 'value', 100)
distribution = messages.annotate(count=Count('id'))
print distribution[0]
# { 'value': 'hello', 'count': 5 }
"""
# Type checking
queryset = find_messages(queryset)
expression = self.get_grouping_expression(queryset, bins=bins, bin_size=bin_size, **kwargs)
if expression is None:
# no results
return queryset.values()
if expression == self.field_name:
# We still have to map back to the requested grouping_key
return MappedValuesQuerySet.create_from(queryset.values(expression), {
expression: grouping_key
})
else:
# Sometimes this step gets complicated
queryset, internal_key = self.select_grouping_expression(queryset,
expression=expression)
# Then use values to group by the grouping key.
queryset = queryset.values(internal_key)
return MappedValuesQuerySet.create_from(queryset, {
internal_key: grouping_key,
})
def group_by(self,
queryset,
grouping_key=None,
values_list=False,
values_list_flat=False,
**kwargs):
"""
Return a ValuesQuerySet that has been grouped by this dimension.
The group value will be available as grouping_key in the dictionaries.
The grouping key defaults to the dimension key.
.. code-block:: python
messages = dim.group_by(messages, 'value')
distribution = messages.annotate(count=Count('id'))
print distribution[0]
# { 'value': 'hello', 'count': 5 }
"""
if grouping_key is None:
grouping_key = self.key
# Type checking
queryset = find_messages(queryset)
# Get the expression that groups this dimension for this queryset
grouping_expression = self.get_grouping_expression(queryset, **kwargs)
queryset, internal_key = self.select_grouping_expression(
queryset, grouping_expression)
# Group the data
queryset = queryset.values(grouping_expression)
if internal_key == grouping_key:
return queryset
else:
# We need to transform the output to match the requested grouping key
# queryset = queryset.extra(select={grouping_key: grouping_expression})
return MappedValuesQuerySet.create_from(
queryset, {internal_key: grouping_key})
def group_by(self, queryset, grouping_key=None, values_list=False, values_list_flat=False, **kwargs):
"""
Return a ValuesQuerySet that has been grouped by this dimension.
The group value will be available as grouping_key in the dictionaries.
The grouping key defaults to the dimension key.
.. code-block:: python
messages = dim.group_by(messages, 'value')
distribution = messages.annotate(count=Count('id'))
print distribution[0]
# { 'value': 'hello', 'count': 5 }
"""
if grouping_key is None:
grouping_key = self.key
# Type checking
queryset = find_messages(queryset)
# Get the expression that groups this dimension for this queryset
grouping_expression = self.get_grouping_expression(queryset, **kwargs)
queryset, internal_key = self.select_grouping_expression(queryset, grouping_expression)
# Group the data
queryset = queryset.values(grouping_expression)
if internal_key == grouping_key:
return queryset
else:
# We need to transform the output to match the requested grouping key
# queryset = queryset.extra(select={grouping_key: grouping_expression})
return MappedValuesQuerySet.create_from(queryset, {
internal_key: grouping_key
})
def render(self,
queryset,
desired_primary_bins=None,
desired_secondary_bins=None):
"""
Given a set of messages (already filtered as necessary),
calculate the data table.
Optionally, a number of primary and secondary bins may be given.
The result is a list of dictionaries. Each
dictionary contains a key for each dimension
and a value key for the count.
"""
if not self.secondary_dimension:
# If there is only one dimension, we should be able to fall back
# on that dimension's group_by() implementation.
queryset = self.primary_dimension.group_by(
queryset,
grouping_key=self.primary_dimension.key,
bins=desired_primary_bins)
return queryset.annotate(value=models.Count('id'))
else:
# Now it gets nasty...
primary_group = self.primary_dimension.get_grouping_expression(
queryset, bins=desired_primary_bins)
secondary_group = self.secondary_dimension.get_grouping_expression(
queryset, bins=desired_secondary_bins)
if primary_group is None or secondary_group is None:
# There is no data to group
return queryset.values()
queryset, internal_primary_key = self.primary_dimension.select_grouping_expression(
queryset, primary_group)
queryset, internal_secondary_key = self.secondary_dimension.select_grouping_expression(
queryset, secondary_group)
# Group the data
queryset = queryset.values(internal_primary_key,
internal_secondary_key)
# Count the messages
queryset = queryset.annotate(value=models.Count('id'))
# We may need to remap some fields
mapping = {}
if internal_primary_key != self.primary_dimension.key:
mapping[internal_primary_key] = self.primary_dimension.key
if internal_secondary_key != self.secondary_dimension.key:
mapping[internal_secondary_key] = self.secondary_dimension.key
if len(mapping) > 0:
return MappedValuesQuerySet.create_from(queryset, mapping)
else:
return queryset