def visualisation():
current_city_name = request.args.get("city_name")
if current_city_name == None:
current_city_name = "Mumbai"
subset = df_iplace[df_iplace['city'] == dropdown.value]
#plot = create_figure(current_city_name)
plot = bc.Bar(subset,
"placename",
values="placename",
agg="count",
plot_width=1000,
plot_height=1000,
title="Popular places")
# Generate the script and HTML for the plot
script, div = components(plot)
# Return the webpage
return """
<!doctype html>
<head>
<title>Popular places</title>
{bokeh_css}
</head>
<body>
<h1>Most Popular Places visited by people!
{div}
{bokeh_js}
{script}
</body>
""".format(script=script,
div=div,
bokeh_css=CDN.render_css(),
bokeh_js=CDN.render_js())
def visuals():
plot = bc.Bar(data=top_ten_states_dataFrame, values='count', label='State')
script, div = components(plot)
error = None
try:
return '''
<!DOCTYPE html>
<html>
<head>
<title>Train Wreck Header</title>
<meta charset="utf-8">
<link rel="stylesheet" type="text/css" href="../static/css/bootstrap.min.css">
<script type="text/javascript" href="../static/js/bootstrap.min.js"></script>
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.2.1/jquery.min.js"></script>
{bokeh_css}
</head>
<header style="background-color: #e6e6e6;">
<!--
<p> started working </p> -->
<nav class="navbar navbar-inverse" style="background-color: #4d4d4d;">
<div class="container-fluid">
<div class="navbar-header">
<a class="navbar-brand" href="/" style="color: white;">Train Wreck Analysis</a>
</div>
<ul class="nav navbar-nav">
<li><a href="/">Home</a></li>
<li><a href="/map/">Map</a></li>
<li><a href="/visuals/">Bar Chart</a></li>
</ul>
</div>
</nav>
</header>
<body>
<div class="container-fluid" style="margin-left: 20px; float:left; border: solid; border-style: groove; border-width: 5px; height: 650px; width: 700px; padding-top: 20px;">
{div}
{bokeh_js}
{script}
</div>
</body>
</html>
'''.format(script=script,
div=div,
bokeh_css=CDN.render_css(),
bokeh_js=CDN.render_js())
except Exception as e:
flash(e)
return render_template('visuals.html', error=error)
def bokeh_bar(df):
"""stolen more or less directly from
http://bokeh.pydata.org/en/0.11.1/docs/user_guide/charts.html
"""
p = bkc.Bar(df,
'animal',
values='number',
title='aminals <3',
tools=BOKEH_TOOLS)
hover = p.select({'type': bkm.HoverTool})
hover.tooltips = """<div>Aminal: @x</div><div>nubmer: @height</div>"""
return bke.components(p)
def states_visualization():
'''
create a dataframe for city and states values
'''
city_State_dataFrame=modifyRow(df2)
'''
finding top ten states with most accidents
'''
states_dataFrame=createStatesList(city_State_dataFrame)
states_dataFrame.apply(lambda x: x.astype(str).str.upper())
states_dataFrame['State'] = states_dataFrame['State'].map(lambda x: x.strip())
#states_dataFrame['States'].str.split(',').str.get(1)
states_dataFrame['State'] = states_dataFrame['State'].apply(lambda x: x.split(',', 1)[-1])
states_dataFrame['State'] = states_dataFrame['State'].map(lambda x: x.strip())
#print states_dataFrame
top_states_dataFrame=states_dataFrame.groupby(['State']).size().reset_index().rename(columns={0:'count'})
#print top_states_dataFrame
all_states_dataFrame=top_states_dataFrame.sort_values(by='count', ascending=False)
# print all_states_dataFrame
all_states=[]
all_states_count=[]
#all_states_tpl=()
for a, b in all_states_dataFrame.itertuples(index=False):
all_states.append(a)
all_states_count.append(b)
#all_states_tpl=all_states_tpl+(a,)
#all_states=zip(all_states,all_states_count)
#all_states=dict(all_states)
#print all_states
top_ten_states_dataFrame=top_states_dataFrame.sort_values(by='count', ascending=False).head(10)
print top_ten_states_dataFrame
#states={}
state=[]
count=[]
states_tpl=()
for a, b in top_ten_states_dataFrame.itertuples(index=False):
state.append(a)
count.append(b)
#states_tpl=states_tpl+(a,)
#states=zip(state,count)
#states=dict(states)
#print states
output_file('bar_graph.html')
plot = bc.Bar(data=top_ten_states_dataFrame,count='count',label='State')
show(plot)
bc.Bar
def learn4():
df1 = ss.read.csv('F:/Research/data/ccFraud.csv',
header=True,
inferSchema=True)
df1.show()
data1 = df1.select('balance').rdd.flatMap(lambda row: row)
print('1:', data1)
data1 = data1.histogram(5)
print('2:', data1)
data = {'bins': data1[0][:-1], 'freq': data1[1]}
print('3:', data)
# Matplotlib绘图
plt1 = plt.figure(figsize=(12, 9))
subplot1 = plt1.add_subplot(2, 2, 1)
subplot1.bar(x=data['bins'], height=data['freq'], width=4000)
subplot1.set_title('balance')
subplot2 = plt1.add_subplot(2, 2, 4)
subplot2.bar(x=data['bins'], height=data['freq'], width=500)
subplot2.set_title('balance')
plt1.show()
# Boken绘图
charts1 = charts.Bar(data,
values='freq',
label='bins',
title='Histogram of \'balance\'')
charts.show(charts1)
# 在性别中各抽取0.02的男女数量，并将抽取数据中['balance', 'numTrans', 'numIntlTrans']三个列提取出来
data_sample1 = df1.sampleBy('gender', {
1: 0.0002,
2: 0.0002
}).select(['balance', 'numTrans', 'numIntlTrans'])
print('0.02%采样后的表：')
data_sample1.show()
# 绘制2D点状图
data_multi = dict([
(elem,
data_sample1.select(elem).rdd.flatMap(lambda row: row).collect())
for elem in ['balance', 'numTrans', 'numIntlTrans']
])
print('点状图表：')
print(len(data_multi), data_multi)
data2 = {data_multi['balance'], data_multi['numTrans']}
charts2 = charts.Scatter(data=data_multi, x='balance', y='numTrans')
charts.show(charts2)
charts2 = charts.Scatter(data=data2, x='balance', y='numTrans')
charts.show(charts2)
def popular_places():
try:
db = mysql.connect()
mycursor = db.cursor()
mycursor.execute("select distinct(city) from placedetails")
cities = []
for row in mycursor:
cities.append(row[0])
form = AdminForm()
form.city.choices = [(city, city) for city in cities]
if request.method == 'POST' and form.validate_on_submit():
cname = request.form['city']
subset = df_iplace[df_iplace['city'] == cname]
places = subset['placename'].value_counts().to_frame()
plot = bc.Bar(places.iloc[0:5, ],
label="index",
values="placename",
plot_width=1000,
plot_height=700,
legend="top_right",
bar_width=0.3,
min_border=30,
xlabel="Places",
ylabel="Count")
script, div = components(plot)
return render_template(
"admin.html",
form=form,
script=script,
title_text="Top 5 Places visited by people in ",
div=div,
bokeh_css=CDN.render_css(),
bokeh_js=CDN.render_js(),
city=cname)
return render_template("admin.html", form=form)
except:
print("Exception occured in admin")
finally:
db.close()
def loadplot(city):
cname = city
subset = df_iplace[df_iplace['city'] == cname]
places = subset['placename'].value_counts().to_frame()
plot = bc.Bar(places.iloc[0:5, ],
label="index",
values="placename",
plot_width=1000,
plot_height=700,
legend="top_right",
bar_width=0.3,
min_border=30,
xlabel="Places",
ylabel="Count")
script, div = components(plot)
return render_template("plot.html",
script=script,
title_text="Top 5 Places visited by people in ",
div=div,
bokeh_css=CDN.render_css(),
bokeh_js=CDN.render_js(),
city=cname)
import pandas as pd
import pymongo
from bokeh import charts
from bokeh.io import push_notebook, show, output_notebook
output_notebook()
client = pymongo.MongoClient()
db = client['hfut']
student_df = pd.DataFrame(list(db['student'].find()),
columns=['学号', '姓名', '性别'])
student_df['入学年份'] = student_df['学号'] // 1000000
p = charts.Bar(student_df, label='入学年份', values='性别', agg='count', stack='性别')
show(p)
# Pclass
survived_pclass = df.groupby('Pclass')['Survived'].value_counts().unstack()
survived_pclass['Rate'] = survived_pclass[1]/(survived_pclass[1] + survived_pclass[0])
survived_pclass
# How does it look graphically?
# In[ ]:
bkp.output_notebook()
bar1 = bkc.Bar(df, values = 'Survived', label = 'Pclass', agg = 'count',
tools='pan,box_zoom,reset,resize,save,hover',
stack=bkc.attributes.cat(columns='Survived', sort=False),
legend='top_left', plot_width=600, plot_height=300)
hover = bar1.select(dict(type = bkm.HoverTool))
hover.tooltips = dict([("Num", "@height{int}")])
bar1.yaxis.axis_label = 'Number of passengers'
bkp.show(bar1)
# Apparently, passengers from higher classes are more likely to survive, both in terms of number and percentage. Furthermore, the survival rates for Pclass 1 and 3 are all quite different from 0.5, hence, 'quite pure'. In another word, during prediction, if I see a passenger is from Pclass 1(3), I would likely to bet he/she will (not) survive. We should take this into account into our future models.
# - *Name*
#
# The next column in the data is 'Name'. What information that we can possibly get from them? Humm.. how about the titles?
# In[ ]: