Result excerpt

Simultaneous posting on the blog: https://leoluistudio.com/blog/42/python%e3%82%92%e4%bd%bf%e3%81%a3%e3%81%a6%e6%9d%b1%e4%ba%ac%e9%83%bd%e5%ae%b6%e8%b3%83%e3%81%ab%e3%81%a4%e3%81%84%e3%81%a6%e3%81%ae%e7%a0%94%e7%a9%b6-4%e3%81%ae3/

Statistical method 2

Programs and results

Calculate median for each building

The reason is that the number of properties in the building is different. For example, if you have 50 properties per building, the weight will be 50 times that of 1 property per building when you calculate the distribution. Take the median for each building so that it is calculated evenly.

conn = sqlite3.connect(‘info.db’)
df = pd.read_sql_query(“SELECT pid,price,area FROM price”, conn)
conn.close()

df[‘municipal’] = df.apply(municipal, axis=1)
dfmedian = df.groupby([‘pid’, ‘municipal’])[‘price’, ‘area’, ‘average’].median()
dfmedian_reset = dfmedian.reset_index(level=’municipal’)

Price distribution by ward (23 wards)

#Graph
fig = go.Figure()
for i in m23_list:
    dfgroup = dfmedian_reset[dfmedian_reset[‘municipal’] == i]
    fig.add_trace(go.Box(x=dfgroup[‘price’], y=dfgroup[‘municipal’], boxpoints=False))
fig.update_traces(orientation=’h’, showlegend=False)
fig.update_xaxes(range=[0, 500000])

Area distribution for each ward (23 wards)

#Graph
fig = go.Figure()
for i in m23_list:
    dfgroup = dfmedian_reset[dfmedian_reset[‘municipal’] == i]
    fig.add_trace(go.Box(x=dfgroup[‘area’], y=dfgroup[‘municipal’], boxpoints=False))
fig.update_traces(orientation=’h’, showlegend=False)
fig.update_xaxes(range=[0, 100])

Calculate the average price (price / area)

def average(df):
    return int(df[‘price’] / df[‘area’])

df[‘average’] = df.apply(average, axis=1)

Distribution of average price for each ward (23 wards)

#Graph
fig = go.Figure()
for i in m23_list:
    dfgroup = dfmedian_reset[dfmedian_reset[‘municipal’] == i]
    fig.add_trace(go.Box(x=dfgroup[‘average’], y=dfgroup[‘municipal’], boxpoints=False))
fig.update_traces(orientation=’h’, showlegend=False)

Relationship between average price and walking time to the station

#Dataframe processing
dfmedian_reset = dfmedian.reset_index(level=’pid’)
dfmedian_reset[‘train’] = dfmedian_reset.apply(trainminute, axis=1)

#Graph
fig = px.scatter(dfmedian_reset, x=’train’, y=’average’, height=500, width=1000)

Relationship between average price and building type

dfmedian_reset = dfmedian.reset_index(level=’pid’)
dfmedian_reset[‘type’] = dfmedian_reset.apply(buildtype, axis=1)
fig = go.Figure()
for i in (['apartment',‘Apartment’]):
    dfgroup = dfmedian_reset[dfmedian_reset[‘type’] == i]
    fig.add_trace(go.Box(x=dfgroup[‘price’], y=dfgroup[‘type’], boxpoints=False))
fig.update_traces(orientation=’h’, showlegend=False)
fig.update_xaxes(range=[0, 500000])
fig.update_layout(height=500, width=1000)