def test_create_model_features(self):
"""
Tests HousePriceModel.create_model_features.
:return: True or False
"""
# Declare and initialize the column name to use for the test, and a
# series of column values.
column_name = 'column'
column_values = pd.Series((1, 2, 3, 4, 5))
# Declare and initialize the source data frame, and give it the test
# column and column values.
source = pd.DataFrame()
source[column_name] = column_values
# Declare and initialize the destination date frame, and copy the
# source column and values.
destination = pd.DataFrame()
HousePriceModel.create_model_feature(destination, source, column_name)
# Get the destination values from the destination, and assert that all
# received values match their expected values.
destination_values = destination[column_name]
return self.assertListEqual(destination_values.tolist(),
column_values.tolist())
def __init__(self, *args, **kwargs):
"""
Initializes the unit test class.
:param args: Argument for unittest.TestCase
:param kwargs: Argument for unittest.TestCase
"""
super(MyTestCase, self).__init__(*args, **kwargs)
self.house_price_model = HousePriceModel()
self.house_price_model.initialize_model()
self.price_accuracy = 100.
def test_create_date(self):
"""
Tests HousePriceModel.create_date.
:return: True or False
"""
# Declare and initialize a test year, month and day.
year = 2017
month = 7
day = 4
# Create the date object, and assert all resulting fields are what we
# expect.
date = HousePriceModel.create_date(year, month, day)
self.assertTrue(date.year == year and date.month == month
and date.day == day)
def test_format_date(self):
"""
Tests HousePriceModel.format_date.
:return: True or False
"""
# Declare and initialize a test year, month and day.
year = 2020
month = 2
day = 29
# Format the date, then parse it.
date = HousePriceModel.format_date(year, month, day)
parsed_date = parse(date)
# Insure that the parsed date matches the expected values.
return self.assertTrue(parsed_date.year == year
and parsed_date.month == month
and parsed_date.day == day)
type bokeh serve --port 5001 main2.py in your terminal
Then you may go to the FirstStop landing page to click the predicting price link
"""
import pandas as pd
from bokeh.io import curdoc
#from bokeh.tile_providers import STAMEN_TONER
from bokeh.models import (GMapPlot, GMapOptions, ColumnDataSource, Circle,
DataRange1d, PanTool, WheelZoomTool, BoxSelectTool,
HoverTool)
from bokeh.layouts import layout
from bokeh.models.widgets import Button, \
Select, Slider, Paragraph, Div
from house_price_model_2 import HousePriceModel
MODEL = HousePriceModel()
MODEL.initialize_model()
LOGO = Div(
text=
"""<img src="https://s3-us-west-2.amazonaws.com/data515logo/logo_title_thinner.PNG"
alt="" />""")
DELIM_1 = Div(text="""<h2><span style="color: #800080;"
width=500 height=15>STEP 1:</span></h2>""")
DELIM_2 = Div(
text=
"""<h2><span style="color: #800080;" width=500 height=15>STEP 2:</span></h2>"""
)
DELIM_3 = Div(
text=
"""<h2><span style="color: #800080;" width=500 height=15>STEP 3:</span></h2>"""
class MyTestCase(unittest.TestCase):
"""
Contains unit tests for the 2nd house price model.
The following methods in HousingPriceModel do not have independent tests
here, and are presumed to be sufficiently tested from calls through the
other unit tests:
1. initialize_model
2. build_model
3. convert_exponential_columns
4. prepare_model_data
5. prepare_test_row
6. read_housing_data
"""
def __init__(self, *args, **kwargs):
"""
Initializes the unit test class.
:param args: Argument for unittest.TestCase
:param kwargs: Argument for unittest.TestCase
"""
super(MyTestCase, self).__init__(*args, **kwargs)
self.house_price_model = HousePriceModel()
self.house_price_model.initialize_model()
self.price_accuracy = 100.
def test_calculate_sales_by_date(self):
"""
Tests HousePriceModel.calculate_sales_day_by_date.
:return: True or False
"""
# Declare and initialize a tuple of test dates.
dates = (self.house_price_model.create_date(2014, 1, 1),
self.house_price_model.create_date(2014, 1, 2),
self.house_price_model.create_date(2014, 1, 31),
self.house_price_model.create_date(2014, 2, 1),
self.house_price_model.create_date(2014, 2, 28),
self.house_price_model.create_date(2014, 3, 1),
self.house_price_model.create_date(2015, 1, 1),
self.house_price_model.create_date(2016, 1, 1),
self.house_price_model.create_date(2016, 2, 29),
self.house_price_model.create_date(2016, 3, 1))
# Declare and initialize the expected results of the test dates
# conversion. Convert the test dates.
expected = (1, 2, 31, 32, 59, 60, 366, 731, 790, 791)
received = self.house_price_model.calculate_sale_day_by_date\
(pd.Series(dates))
# Assert that all received test dates match their expected values.
return self.assertListEqual(received.tolist(), list(expected))
def test_calculate_sales_by_day(self):
"""
Tests HousePriceModel.calculate_sales_day_by_day.
:return: True or False
"""
return self.assertTrue(
self.house_price_model.calculate_sale_day_by_day(2016, 2, 29) ==
790)
def test_can_predict(self):
"""
Tests HousePriceModel.can_predict.
:return: True or False
"""
return self.assertTrue(self.house_price_model.can_predict)
def test_create_date(self):
"""
Tests HousePriceModel.create_date.
:return: True or False
"""
# Declare and initialize a test year, month and day.
year = 2017
month = 7
day = 4
# Create the date object, and assert all resulting fields are what we
# expect.
date = HousePriceModel.create_date(year, month, day)
self.assertTrue(date.year == year and date.month == month
and date.day == day)
def test_create_model_features(self):
"""
Tests HousePriceModel.create_model_features.
:return: True or False
"""
# Declare and initialize the column name to use for the test, and a
# series of column values.
column_name = 'column'
column_values = pd.Series((1, 2, 3, 4, 5))
# Declare and initialize the source data frame, and give it the test
# column and column values.
source = pd.DataFrame()
source[column_name] = column_values
# Declare and initialize the destination date frame, and copy the
# source column and values.
destination = pd.DataFrame()
HousePriceModel.create_model_feature(destination, source, column_name)
# Get the destination values from the destination, and assert that all
# received values match their expected values.
destination_values = destination[column_name]
return self.assertListEqual(destination_values.tolist(),
column_values.tolist())
def test_format_date(self):
"""
Tests HousePriceModel.format_date.
:return: True or False
"""
# Declare and initialize a test year, month and day.
year = 2020
month = 2
day = 29
# Format the date, then parse it.
date = HousePriceModel.format_date(year, month, day)
parsed_date = parse(date)
# Insure that the parsed date matches the expected values.
return self.assertTrue(parsed_date.year == year
and parsed_date.month == month
and parsed_date.day == day)
def test_get_base_date(self):
"""
Tests HousePriceModel.get_base_date.
:return: True or False
"""
# Declare and initialize a test year, month and day.
year = 2014
month = 1
day = 1
# Insure that the base date matches the expected values.
base_date = self.house_price_model.get_base_date()
return self.assertTrue(base_date.year == year
and base_date.month == month
and base_date.day == day)
def test_get_day_offset(self):
"""
Tests HousePriceModel.get_day_offset.
:return: True of False
"""
self.assertTrue(HousePriceModel.get_day_offset() == 1)
def test_get_mean_response(self):
"""
Tests HousePriceModel.get_mean_response.
:return: True or False
"""
return self.assertAlmostEqual(
self.house_price_model.get_mean_response(),
537798.,
delta=self.price_accuracy)
def test_get_model(self):
"""
Tests HousePriceModel.get_model.
:return: True or False
"""
# Test whether the model is a RidgeCV model, and whether
# the coefficients from the model match.
model = self.house_price_model.get_model()
return self.assertTrue(isinstance(model, RidgeCV)) \
and self.assertListEqual(list(model.coef_),
list(self.house_price_model.get_model_coefficients()))
def test_get_model_coefficients(self):
"""
Tests HousePriceModel.get_model.
:return: True or False
"""
# Just run the same test as the one for the model itself.
return self.test_get_model()
def test_get_predictors(self):
"""
Tests HousePriceModel.get_predictors.
:return: True or False
"""
return self.assertTrue(
self.house_price_model.get_predictors().shape[1] == 9)
def test_get_sales_data(self):
"""
Tests HousePriceModel.get_sales_data.
:return: True or False
"""
return self.assertTrue(
len(self.house_price_model.get_sales_data().columns) == 22)
def test_get_scaler(self):
"""
Tests HousePriceModel.get_scaler.
:return: True or False
"""
return self.assertTrue(
isinstance(self.house_price_model.get_scaler(), StandardScaler))
def test_get_zip_code_dict(self):
"""
Tests HousePriceModel.get_zip_code_dict.
:return: True or False
"""
return self.assertTrue(
isinstance(self.house_price_model.get_zip_code_dict(), dict))
def test_lookup_zipcode_by_number(self):
"""
Tests HousePriceModel.lookup_zip_code_by_number.
:return: True or False
"""
# Declare and initialize the zip codes tuple.
zip_codes = (98002, 98168, 98032, 98001, 98148, 98023, 98188, 98003,
98030, 98031, 98198, 98055, 98178, 98042, 98022, 98106,
98092, 98058, 98108, 98146, 98038, 98133, 98118, 98056,
98155, 98126, 98019, 98014, 98028, 98166, 98125, 98070,
98011, 98059, 98034, 98065, 98136, 98072, 98117, 98107,
98103, 98144, 98029, 98027, 98007, 98116, 98115, 98122,
98052, 98008, 98177, 98053, 98077, 98074, 98075, 98199,
98033, 98005, 98119, 98006, 98105, 98109, 98102, 98112,
98040, 98004)
# Declare and initialize the locations tuple.
locations = (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30,
31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44,
45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58,
59, 60, 61, 62, 63, 64, 65)
# Assert that lookup_zipcode_by_number returns the right value for
# each zip code.
return self.assertListEqual([
self.house_price_model.look_up_zipcode_by_number(zipcode)
for zipcode in zip_codes
], list(locations))
def test_lookup_zipcode_by_string(self):
"""
Tests HousePriceModel.lookup_zip_code_by_string.
:return: True or False
"""
# Declare and initialize the zip codes tuple.
zip_codes = ('98002', '98168', '98032', '98001', '98148', '98023',
'98188', '98003', '98030', '98031', '98198', '98055',
'98178', '98042', '98022', '98106', '98092', '98058',
'98108', '98146', '98038', '98133', '98118', '98056',
'98155', '98126', '98019', '98014', '98028', '98166',
'98125', '98070', '98011', '98059', '98034', '98065',
'98136', '98072', '98117', '98107', '98103', '98144',
'98029', '98027', '98007', '98116', '98115', '98122',
'98052', '98008', '98177', '98053', '98077', '98074',
'98075', '98199', '98033', '98005', '98119', '98006',
'98105', '98109', '98102', '98112', '98040', '98004')
# Declare and initialize the locations tuple.
locations = (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30,
31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44,
45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58,
59, 60, 61, 62, 63, 64, 65)
# Assert that lookup_zipcode_by_string returns the right value for
# each zip code.
return self.assertListEqual([
self.house_price_model.look_up_zipcode_by_string(zipcode)
for zipcode in zip_codes
], list(locations))
def test_prediction_one(self):
"""
Performs a test house price prediction for 1817 N 51st St,
Seattle, WA, 98103.
:return: True or False
"""
# Declare and initialize a test row.
features = {
'sale_day':
self.house_price_model.calculate_sale_day_by_day(2017, 7, 1),
'bathrooms':
2.5,
'sqft_living':
1430,
'sqft_lot':
3210,
'waterfront':
0,
'view':
0,
'condition':
5,
'grade':
6,
'location':
self.house_price_model.look_up_zipcode_by_string('98103')
}
# Make a prediction and check it.
return self.assertAlmostEqual(self.house_price_model.predict(features),
625696.,
delta=self.price_accuracy)
def test_prediction_two(self):
"""
Performs a test house price prediction for 5218 Greenwood Ave N,
Seattle, WA, 98103.
:return: True or False
"""
# Declare and initialize a test row.
features = {
'sale_day':
self.house_price_model.calculate_sale_day_by_day(2017, 7, 1),
'bathrooms':
3,
'sqft_living':
2640,
'sqft_lot':
3920,
'waterfront':
0,
'view':
0,
'condition':
5,
'grade':
6,
'location':
self.house_price_model.look_up_zipcode_by_string('98103')
}
# Make a prediction and check it.
return self.assertAlmostEqual(self.house_price_model.predict(features),
775624.,
delta=self.price_accuracy)
def test_prediction_three(self):
"""
Performs a test house price prediction for 216 Pike St SE,
Auburn, WA 98002
:return: True or False
"""
# Declare and initialize a test row.
features = {
'sale_day':
self.house_price_model.calculate_sale_day_by_day(2017, 7, 1),
'bathrooms':
1,
'sqft_living':
1060,
'sqft_lot':
8000,
'waterfront':
0,
'view':
0,
'condition':
5,
'grade':
6,
'location':
self.house_price_model.look_up_zipcode_by_string('98002')
}
# Make a prediction and check it.
return self.assertAlmostEqual(self.house_price_model.predict(features),
252879.,
delta=self.price_accuracy)
def test_get_day_offset(self):
"""
Tests HousePriceModel.get_day_offset.
:return: True of False
"""
self.assertTrue(HousePriceModel.get_day_offset() == 1)