def index(building):
buildings = building_list()
flag = session.get('flag')
url = session.get('url')
session['flag'] = False
session['url'] = ''
print(building)
if '*' in building:
Html_file = open(
os.path.join(mydir, 'templates/building/' + building + '.html'),
"r")
image = Html_file.read()
x_coord, y_coord, screen_xs, screen_ys, zoomLevels, length = fetch_data(
db, building.lower(), placeholder=None)
building_string = 'building/template.html'
return render_template(building_string,
x_coord=x_coord,
y_coord=y_coord,
screen_xs=screen_xs,
screen_ys=screen_ys,
zoomLevels=zoomLevels,
length=length,
building=building,
image=image,
buildings=buildings,
flag=flag,
url=url)
else:
x_coord, y_coord, screen_xs, screen_ys, zoomLevels, length = fetch_data(
db, building.lower(), placeholder=None)
building_string = 'building/' + building.lower() + '.html'
return render_template(building_string,
x_coord=x_coord,
y_coord=y_coord,
screen_xs=screen_xs,
screen_ys=screen_ys,
zoomLevels=zoomLevels,
length=length,
buildings=buildings,
flag=flag,
url=url)
from sklearn.datasets import fetch_20newsgroups
import helper as hlp
import taskd as td
from sklearn.linear_model import LogisticRegression
from sklearn import metrics
import sklearn.metrics as smet
from sklearn.metrics import roc_curve
import matplotlib.pyplot as plt
categories = hlp.fetch_categories()
twenty_train, twenty_test = hlp.fetch_data()
hlp.classify_into_two_class(twenty_train)
hlp.classify_into_two_class(twenty_test)
svdListTrain = td.getsvdListTrain()
nmfListTrain = td.getnmfListTrain()
svdListTest = td.getsvdListTest()
nmfListTest = td.getnmfListTest()
classifier = LogisticRegression(C=10000)
def classifyLR(train, test):
classifier.fit(train, twenty_train.target)
predicted = classifier.predict(test)
predicted_probs = classifier.predict_proba(test)
hlp.getStats(twenty_test.target, predicted)
hlp.plot_roc(twenty_test.target, predicted_probs[:, 1],
'Logistic Regression')
for min_df in [2, 5]:
import helper as hlp import task1 as t1 dataset = hlp.fetch_data() hlp.classify_into_two_class(dataset) labels = hlp.fetch_labels(dataset) tfidf_matrix = t1.getTFIDF_matrix(dataset, 3) km = hlp.getKmeans(2) km.fit(tfidf_matrix) hlp.getStats(labels, km.labels_)
def main(message):
"""
Background:
This app fetches bike share data from the City of Toronto and the user's location from the client. It then determines the top 5 closest bike stations to the user and returns a route to the closest station prior to sending this information back to the client to be plotted on a map.
Input:
message: GEO data that comes from the client containing the user's longitude and latitude.
Output:
client_data: Socket_io sends (emits) data back to the client containing data from the top 5 closest stations to the user and
any route information. This data will be used to plot the top 5 stations on a leaflet map.
"""
# Fetch 2 toronto bike share data feeds and store in stn_attr, stn_status
stn_attr, stn_status = fetch_data()
# Merge Station Information and Station Status
all_stn_data = join_stn_data(stn_status, stn_attr)
# Store User's Lat / Lon data from the client via socketio into a data variable
# print(message)
data = np.array(list(message.values()))
# The user's lat / lon coordinates
mylat = data[0][0]
mylon = data[0][1]
# mycoord will be used in some functions from helper.py
mycoord = [mylat,mylon]
# get a list of stations and their distances sorted by closest station to the user's lat / lon
distances_to_all = get_closest_stn(mycoord, all_stn_data)
# Get the closest stations lat / lon and the user's distance to the closest station (in km)
closest_lat = distances_to_all[0][4]
closest_lon = distances_to_all[0][5]
closest_distance = distances_to_all[0][1]
# If the user's distance is less than 2 km away from a station,
# fetch all of the lat/lon positions between the user and the closest station.
if closest_distance <= 2:
# get route lat / lon data
route = plot_route(mylat,mylon,closest_lat,closest_lon)
if route == None:
route = 0
print('No route data available')
else:
route
print(route)
else:
# do nothing
route = 0
print("Route too far to plot")
# slice distances_to_all and get the top 5 closest stations
top5closest_stns = distances_to_all[0:5]
# Prepare to send the top 5 closest stations and route information back to the client
client_data = [top5closest_stns, route]
#print(client_data)
# Send the top 5 closest stations and route information back to the client
emit('my_response', {'data': client_data})
