def calculate_flow(self, atomic_params):
"""
checking if the shelf is sorted like the brands list.
:return: 100 if it's fine, 0 otherwise.
"""
population_entity_type = Converters.convert_type(atomic_params[Const.ENTITY_TYPE])
progression_list = map(lambda x: x.strip(), atomic_params[Const.ENTITY_VAL].split(','))
location_entity_type = Converters.convert_type(atomic_params[Const.TYPE_FILTER])
location_values = map(lambda x: x.strip(), atomic_params[Const.VALUE_FILTER].split(','))
is_in_param = False if atomic_params[Const.IN_NOT_IN] == u'Not in' else True
filter_loc_param = self.scif[location_entity_type].isin(location_values) if is_in_param else\
~self.scif[location_entity_type].isin(location_values)
filtered_scif = self.scif[
filter_loc_param &
(self.scif['tagged'] >= 1) &
(self.scif[population_entity_type].isin(progression_list))]
join_on = ['scene_fk', 'product_fk']
match_product_join_scif = pd.merge(filtered_scif, self.match_product_in_scene, on=join_on, how='left',
suffixes=('_x', '_matches'))
progression_field = 'brand_name'
group_column = 'scene_fk'
progression_cross_shelves_true = (self.tool_box_for_flow.progression(
df=match_product_join_scif, progression_list=progression_list, progression_field=progression_field,
at_least_one=False, left_to_right=False, cross_bays=True, cross_shelves=True,
include_stacking=False, group_by=group_column))
progression_cross_shelves_false = (self.tool_box_for_flow.progression(
df=match_product_join_scif, progression_list=progression_list, progression_field=progression_field,
at_least_one=False, left_to_right=False, cross_bays=True, cross_shelves=False,
include_stacking=False, group_by=group_column))
return 100.0 * (progression_cross_shelves_true or
progression_cross_shelves_false)
def calculate_sos(self, atomic_params):
"""
:param atomic_params: dict - atomic kpi line from the template
:return: the percent of SOS (if it's binary - 100 if more than target, otherwise 0).
"""
numerator_type = atomic_params[Const.ENTITY_TYPE_NUMERATOR]
numerator_value = atomic_params[Const.NUMERATOR]
denominator_type = atomic_params[Const.ENTITY_TYPE_DENOMINATOR]
denominator_value = atomic_params[Const.DENOMINATOR]
in_or_not = atomic_params[Const.IN_NOT_IN]
filter_type = Converters.convert_type(atomic_params[Const.TYPE_FILTER])
filter_value = atomic_params[Const.VALUE_FILTER]
denominator_filters = self.get_default_filters(denominator_type, denominator_value)
numerator_filters = self.get_default_filters(numerator_type, numerator_value)
if in_or_not:
numerator_filters = self.update_filters(numerator_filters, in_or_not, filter_type, filter_value)
denominator_filters = self.update_filters(denominator_filters, in_or_not, filter_type, filter_value)
atomic_score = self.tools.calculate_share_of_shelf(
sos_filters=numerator_filters, **denominator_filters) * 100
if atomic_params[Const.SCORE] == Const.BINARY:
try:
return 100 * (atomic_score >= float(atomic_params[Const.targets_line][
self.store_type]))
except ValueError:
Log.warning('The target for {} is bad in store {}'.format(
atomic_params[Const.ATOMIC_NAME], self.store_type))
return 0.0
elif atomic_params[Const.SCORE] != Const.NUMERIC:
Log.error('The score is not numeric and not binary.')
return atomic_score
def calculate_scene_count(self, atomic_params):
"""
:param atomic_params: dict - atomic kpi line from the template
:return: int - amount of scenes.
"""
filters = {Converters.convert_type(
atomic_params[Const.ENTITY_TYPE]): map(lambda x: x.strip(), atomic_params[Const.ENTITY_VAL].split(','))}
scene_count = self.tools.calculate_number_of_scenes(**filters)
return scene_count
def get_default_filters(type_name, value_name):
"""
:param type_name: string that contains list of types
:param value_name: string that contains list of values in the same length
:return: filter as dict.
"""
if ',' in type_name:
types = type_name.split(',')
types = map(lambda x: x.strip(), types)
values = value_name.split(',')
values = map(lambda x: x.strip(), values)
filters = {}
if len(types) != len(values):
Log.warning('there are {} types and {} values, should be the same amount'.format(
len(types), len(values)))
else:
for i in xrange(len(types)):
filters[Converters.convert_type(types[i])] = values[i]
else:
filters = {Converters.convert_type(type_name): map(lambda x: x.strip(), value_name.split(','))}
return filters
def calculate_all_availability(self, type1, value1, type2, value2, in_or_not, filter_type, filter_value):
"""
:param atomic_params: dict - atomic kpi line from the template.
checks the kind of the survey and sends it to the match function
:return: float - score.
"""
if not type1 or not value1:
Log.warning('There is no type and value in the atomic availability')
return 0.0
type1 = Converters.convert_type(type1)
value1 = value1.split(',')
value1 = map(lambda x: x.strip(), value1)
type2 = Converters.convert_type(type2)
value2 = value2
filters = {type1: value1}
if type2 and value2:
value2 = value2.split(',')
value2 = map(lambda x: x.strip(), value2)
filters[type2] = value2
if in_or_not:
filters = self.update_filters(filters, in_or_not, filter_type, filter_value)
return self.tools.calculate_availability(**filters)
def calculate_availability(self, atomic_params):
"""
:param atomic_params: dict - atomic kpi line from the template
:return: 100 if this product is available, 0 otherwise.
"""
availability = self.calculate_all_availability(atomic_params[Const.ENTITY_TYPE],
atomic_params[Const.ENTITY_VAL],
atomic_params[Const.ENTITY_TYPE2],
atomic_params[Const.ENTITY_VAL2],
atomic_params[Const.IN_NOT_IN],
Converters.convert_type(atomic_params[Const.TYPE_FILTER]),
atomic_params[Const.VALUE_FILTER])
return 100.0 * (availability > 0)
def update_filters(self, filters, in_or_not, filter_type, filter_value):
"""
:param filters: the source filters as dict.
:param in_or_not: boolean if it should include or not include
:param filter_type: str
:param filter_value: str
adds to exist filter if to include/exclude one more condition.
:return: dict - the updated filter.
"""
filter_type = Converters.convert_type(filter_type)
if "Not" in in_or_not:
list_of_negative = list(self.scif[~(self.scif[filter_type] == filter_value)][filter_type].unique())
filters[filter_type] = list_of_negative
elif "In" in in_or_not:
filters[filter_type] = filter_value
else:
Log.warning('The value in "In/Not In" in the template should be "Not in", "In" or empty')
return filters
def calculate_planogram_new(self, atomic_params):
"""
:param atomic_params: dict - atomic kpi line from the template
:return: 100 if there is scene which has at least one correctly positioned product, 0 otherwise.
"""
type_name = Converters.convert_type(atomic_params[Const.ENTITY_TYPE])
values = map(lambda x: x.strip(), atomic_params[Const.ENTITY_VAL].split(','))
wanted_answer = float(atomic_params[Const.ACCEPTED_ANSWER_RESULT])
filtered_scenes = self.scif[self.scif[type_name].isin(values)][ScifConsts.SCENE_FK].unique()
scenes_passing = 0
for scene in filtered_scenes:
incor_tags = self.match_product_in_scene[(self.match_product_in_scene[ScifConsts.SCENE_FK]
== scene) &
(~(self.match_product_in_scene[MatchesConsts.COMPLIANCE_STATUS_FK]
== 3))]
if len(incor_tags) == 0:
scenes_passing += 1
score = 100 if scenes_passing >= wanted_answer else 0
return score
def calculate_planogram(self, atomic_params):
"""
:param atomic_params: dict - atomic kpi line from the template
:return: 100 if there is planogram, 0 otherwise.
"""
type_name = Converters.convert_type(atomic_params[Const.ENTITY_TYPE])
values = atomic_params[Const.ENTITY_VAL].split(',')
values = map(lambda x: x.strip(), values)
wanted_answer = float(atomic_params[Const.ACCEPTED_ANSWER_RESULT])
filtered_scenes = self.scif[self.scif[type_name].isin(values)]['scene_id'].unique()
count = 0
for scene_id in filtered_scenes:
query = CCZAQueries.getPlanogramByTemplateName(scene_id)
planogram = pd.read_sql_query(query, self.rds_conn.db)
if 1 in planogram['match_compliance_status'].unique().tolist():
count += 1
if count >= wanted_answer:
return 100.0
return 0.0
