class FilterSelection(object):
"""Class for Selecting Programs Matching Filters."""
def __init__(self,
overall_rank=None,
usnews=None,
cwur=None,
forbes=None,
times=None,
state=None,
city=None,
zipcode=None,
region=None,
fees_in_state=None,
fees_out_of_state=None,
gpa=None,
verbal=None,
quant=None,
boarding=None,
books=None,
overall_expenses=None,
admission_rate=None,
areas_of_interest=None):
"""Initializing Filter Class."""
self.overall_rank = overall_rank
self.usnews = usnews
self.cwur = cwur
self.forbes = forbes
self.times = times
self.state = state
self.city = city
self.zipcode = zipcode
self.region = region
self.fees_in_state = fees_in_state
self.fees_out_of_state = fees_out_of_state
self.gpa = gpa
self.verbal = verbal
self.quant = quant
self.boarding = boarding
self.books = books
self.overall_expenses = overall_expenses
self.admission_rate = admission_rate
self.areas_of_interest = areas_of_interest
# List of programs matching atleast one of the requested criteria
self.unique_programs = None
# List of programs meeting all requested criteria
self.common_programs = []
# Verbose program dict with all matching info
self.matches = None
# creating the object of Filters class
self.filters = Filters()
# Program Count
self.count = self.filters.count
def filter_programs(self):
"""Filter the programs."""
# Deploy Filter Selection
self.run_selected_filters()
# Getting all the unique programs
self.unique_programs = list(self.matches.keys())
# Getting all common program intersection
common = set.intersection(*map(set, self.common_programs))
# Validating we atleast have one program matching one criteria
# Fall back to all programs if no match is found
if len(self.unique_programs) < 1:
# Setting unique programs to all programs
self.unique_programs = []
self.unique_programs.extend(range(0, self.count - 1))
# Setting all matches to zero
self.matches = []
self.matches = dict.fromkeys(self.unique_programs, 1)
# Setting all common programs to all programs
common = []
common = common.extend(range(0, self.count - 1))
return (common, self.unique_programs, self.matches)
def update_results(self, matching_programs):
"""Helper method to update matching programs results."""
self.common_programs.append(matching_programs)
for program in matching_programs:
# Update the verbose program dict of results
if self.matches is None:
self.matches = {}
self.matches[program] = 1
elif program in self.matches:
self.matches[program] += 1
else:
self.matches[program] = 1
def run_selected_filters(self):
"""Running selected filters."""
if self.overall_rank is not None:
matching_programs = self.filters.filter_rank_overall(
self.overall_rank)
self.update_results(matching_programs)
if self.usnews is not None:
matching_programs = self.filters.filter_rank_usnews(self.usnews)
self.update_results(matching_programs)
if self.cwur is not None:
matching_programs = self.filters.filter_rank_cwur(self.cwur)
self.update_results(matching_programs)
if self.forbes is not None:
matching_programs = self.filters.filter_rank_forbes(self.forbes)
self.update_results(matching_programs)
if self.times is not None:
matching_programs = self.filters.filter_rank_times(self.times)
self.update_results(matching_programs)
if self.state is not None:
matching_programs = self.filters.filter_location_state(self.state)
self.update_results(matching_programs)
if self.city is not None:
matching_programs = self.filters.filter_location_city(self.city)
self.update_results(matching_programs)
if self.zipcode is not None:
matching_programs = self.filters.filter_location_zip(self.zipcode)
self.update_results(matching_programs)
if self.region is not None:
matching_programs = self.filters.filter_location_region(
self.region)
self.update_results(matching_programs)
if self.fees_in_state is not None:
matching_programs = self.filters.filter_fees_in_state(
self.fees_in_state)
self.update_results(matching_programs)
if self.fees_out_of_state is not None:
matching_programs = self.filters.filter_fees_out_state(
self.fees_out_of_state)
self.update_results(matching_programs)
if self.gpa is not None:
matching_programs = self.filters.filter_gpa(self.gpa)
self.update_results(matching_programs)
if self.verbal is not None:
matching_programs = self.filters.filter_gre_verbal(self.verbal)
self.update_results(matching_programs)
if self.quant is not None:
matching_programs = self.filters.filter_gre_quant(self.quant)
self.update_results(matching_programs)
if self.boarding is not None:
matching_programs = self.filters.filter_boarding(self.boarding)
self.update_results(matching_programs)
if self.books is not None:
matching_programs = self.filters.filter_books(self.books)
self.update_results(matching_programs)
if self.overall_expenses is not None:
matching_programs = self.filters.filter_overall_expenses(
self.overall_expenses)
self.update_results(matching_programs)
if self.admission_rate is not None:
matching_programs = self.filters.filter_admission_rate(
self.admission_rate)
self.update_results(matching_programs)
class Recommend(object):
"""Class for Selecting Programs Matching Filters."""
def __init__(self,
overall_rank=None,
rank=None,
usnews=None,
cwur=None,
forbes=None,
times=None,
state=None,
city=None,
zipcode=None,
region=None,
budget=None,
location=None,
in_state=None,
out_of_state=None,
gpa=None,
verbal=None,
quant=None,
boarding=None,
books=None,
overall_expenses=None,
admission_rate=None,
aoi=None):
"""Initializing Filter Class."""
self.rank = rank
self.overall_rank = overall_rank
self.usnews = usnews
self.cwur = cwur
self.forbes = forbes
self.times = times
self.state = state
self.city = city
self.zipcode = zipcode
self.region = region
self.location = location
self.in_state = in_state
self.out_of_state = out_of_state
self.gpa = gpa
self.verbal = verbal
self.quant = quant
self.budget = budget
self.boarding = boarding
self.books = books
self.overall_expenses = overall_expenses
self.admission_rate = admission_rate
self.aoi = aoi
# List of programs matching atleast one of the requested criteria
self.unique_programs = None
# List of programs meeting all requested criteria
self.common_programs = []
# Verbose program dict with all matching info
self.matches = None
# creating the object of Filters class
self.filters = Filters()
# Program Count
self.count = self.filters.count
self.firebase = Firebase()
def filter_programs(self):
"""Filter the programs."""
# Deploy Filter Selection
specializations, user_selected_criteria = self.run_selected_filters()
# Getting all the unique programs
self.unique_programs = list(self.matches.keys())
# Getting all common program intersection
common = set.intersection(*map(set, self.common_programs))
# Validating we atleast have one program matching one criteria
# Fall back to all programs if no match is found
if len(self.unique_programs) < 1:
# Setting unique programs to all programs
self.unique_programs = []
self.unique_programs.extend(range(0, self.count - 1))
# Setting all matches to zero
self.matches = []
self.matches = dict.fromkeys(self.unique_programs, 1)
# Setting all common programs to all programs
common = []
common = common.extend(range(0, self.count - 1))
return (common, self.unique_programs, self.matches, specializations,
user_selected_criteria)
def update_results(self, matching_programs):
"""Helper method to update matching programs results."""
self.common_programs.append(matching_programs)
for program in matching_programs:
# Update the verbose program dict of results
if self.matches is None:
self.matches = {}
self.matches[program] = 1
elif program in self.matches:
self.matches[program] += 1
else:
self.matches[program] = 1
def run_selected_filters(self):
specializations = []
user_selected_criteria = []
#Based on the UI Options
if self.rank is not None:
matching_programs = self.filters.filter_rank_usnews(self.rank)
self.update_results(matching_programs)
matching_programs = self.filters.filter_rank_cwur(self.rank)
self.update_results(matching_programs)
matching_programs = self.filters.filter_rank_forbes(self.rank)
self.update_results(matching_programs)
matching_programs = self.filters.filter_rank_times(self.rank)
self.update_results(matching_programs)
rank_list = ["usnews", "cwur", "forbes", "times"]
user_selected_criteria.extend(rank_list)
if self.budget is not None:
matching_programs = self.filters.filter_fees_out_state(self.budget)
self.update_results(matching_programs)
matching_programs = self.filters.filter_fees_in_state(self.budget)
self.update_results(matching_programs)
matching_programs = self.filters.filter_boarding(self.budget)
self.update_results(matching_programs)
matching_programs = self.filters.filter_books(self.budget)
self.update_results(matching_programs)
user_selected_criteria.append("in_state")
user_selected_criteria.append("out_of_state")
user_selected_criteria.append("boarding")
user_selected_criteria.append("books")
if self.aoi is not None:
specializations.extend(self.aoi)
if self.location is not None:
matching_programs = self.filters.filter_location_region(
self.location)
self.update_results(matching_programs)
"""Individual filters."""
if self.overall_rank is not None:
matching_programs = self.filters.filter_rank_overall(
self.overall_rank)
self.update_results(matching_programs)
if self.usnews is not None:
matching_programs = self.filters.filter_rank_usnews(self.usnews)
self.update_results(matching_programs)
if self.cwur is not None:
matching_programs = self.filters.filter_rank_cwur(self.cwur)
self.update_results(matching_programs)
if self.forbes is not None:
matching_programs = self.filters.filter_rank_forbes(self.forbes)
self.update_results(matching_programs)
if self.times is not None:
matching_programs = self.filters.filter_rank_times(self.times)
self.update_results(matching_programs)
if self.state is not None:
matching_programs = self.filters.filter_location_state(self.state)
self.update_results(matching_programs)
if self.city is not None:
matching_programs = self.filters.filter_location_city(self.city)
self.update_results(matching_programs)
if self.zipcode is not None:
matching_programs = self.filters.filter_location_zip(self.zipcode)
self.update_results(matching_programs)
if self.in_state is not None:
matching_programs = self.filters.filter_fees_in_state(
self.in_state)
self.update_results(matching_programs)
if self.out_of_state is not None:
matching_programs = self.filters.filter_fees_out_state(
self.out_of_state)
self.update_results(matching_programs)
if self.gpa is not None:
matching_programs = self.filters.filter_gpa(self.gpa)
self.update_results(matching_programs)
if self.verbal is not None:
matching_programs = self.filters.filter_gre_verbal(self.verbal)
self.update_results(matching_programs)
if self.quant is not None:
matching_programs = self.filters.filter_gre_quant(self.quant)
self.update_results(matching_programs)
if self.boarding is not None:
matching_programs = self.filters.filter_boarding(self.boarding)
self.update_results(matching_programs)
if self.books is not None:
matching_programs = self.filters.filter_books(self.books)
self.update_results(matching_programs)
if self.overall_expenses is not None:
matching_programs = self.filters.filter_overall_expenses(
self.overall_expenses)
self.update_results(matching_programs)
if self.admission_rate is not None:
matching_programs = self.filters.filter_admission_rate(
self.admission_rate)
self.update_results(matching_programs)
return specializations, user_selected_criteria
def get_user_criteria(self):
if self.rank is not None:
self.user_selected_criteria.extend("usnews")
self.user_selected_criteria.extend("forbes")
self.user_selected_criteria.extend("times")
self.user_selected_criteria.extend("cwur")
if self.in_state is not None:
self.user_selected_criteria.extend("in_state")
self.user_selected_criteria.extend("out_of_state")
self.user_selected_criteria.extend("boarding")
self.user_selected_criteria.extend("books")
self.user_selected_criteria.extend("admission_rate")
if self.gpa is not None:
self.user_selected_criteria.extend("gpa")
if self.verbal is not None:
self.user_selected_criteria.extend("verbal")
if self.quant is not None:
self.user_selected_criteria.extend("quant")
if self.aoi is not None:
self.specializations_list.extend(aoi)
else:
self.specializations_list = [
'information_assurance_cyber_security',
'business_intelligence', 'computer_networks',
'web_application_development', 'library_science',
'management_consulting', 'human_center_design_engineering',
'information_architecture', 'software_engineering',
'data_science_analytics', 'distributed_systems'
]
return (self.user_selected_criteria, self.specializations_list)
#Getting the json list after the filtering
def get_filtered_json(self, program_list):
json_list = []
for each in program_list:
result = self.firebase.get_detailed_program(each)
json_list.append(result)
return json_list
def construct_dataframe(self, json_list, program_list):
df = pd.io.json.json_normalize(json_list)
df.columns = df.columns.map(lambda x: x.split(".")[-1])
c_list = list(df.columns.values)
col_list = []
for each in c_list:
each = each.encode("utf-8")
col_list.append(each)
df = df.loc[:, [
'gpa', 'quant', 'verbal', 'admission_rate',
'business_intelligence', 'computer_networks',
'data_science_analytics', 'distributed_systems',
'human_center_design_engineering', 'information_architecture',
'information_assurance_cyber_security', 'library_science',
'management_consulting', 'software_engineering',
'web_application_development', 'in_state', 'out_of_state',
'boarding', 'books', 'other', 'length', 'cwur', 'forbes', 'times',
'usnews'
]]
df["program_id"] = program_list
return df
def recommend_programs(self):
set_list, program_list, program_dict, specializations, user_selected_criteria = self.filter_programs(
)
json_list = self.get_filtered_json(program_list)
df = self.construct_dataframe(json_list, program_list)
#user_selected_criteria=["forbes","out_of_state"]
#specializations_list=['information_assurance_cyber_security',
# 'business_intelligence',
# 'computer_networks']
#user_selected_criteria, specializations_list=self.get_user_criteria()
match = MatchingAlgo(df, program_list, specializations,
user_selected_criteria)
recommendations = match.rank_programs()
return recommendations
