#!/usr/bin/env python
# coding: utf-8

# In[96]:


# Import quandl to retrieve data
# Pandas and matplotlib standard imports for data analysis
import quandl
import pandas as pd
import matplotlib.pyplot as plt
get_ipython().run_line_magic('matplotlib', 'inline')

# Use Plotly for better graphs
import plotly.offline as plotly
import plotly.graph_objs as go

# Offline Plotting
plotly.init_notebook_mode(connected=True)


# In[97]:


# Constants

# Populate your own quandl API Token
# Populate your own mapbox API Token
quandl.ApiConfig.api_key = {YOUR API KEY}
mapbox_access_token = {YOUR API KEY}

# Start date and end date should be the same
start_date = '2018-06-30'
end_date = '2018-06-30'

# Specify Zillow Indicator we want to graph
indicator = 'MRPFAH'

# Specify the county and state we want to generate the map for
county = 'Alameda'
state = 'CA'


# In[98]:


# We are going to fetch the data for each zip code in the county and add that to a single dataframe
def get_city_data(cities, indicator):
    df_array = []
    for city, code in cities.items():
        quandl_db_code = f'ZILLOW/Z{code}_{indicator}'
        try:
            data = quandl.get(quandl_db_code, start_date=start_date, end_date=end_date)
            data.rename(index=str, columns={'Value': city}, inplace=True)
            data.index = pd.to_datetime(data.index, format='%Y/%m/%d')
            df_array.append(data)
        except:
            print(f'Error fetching for: {quandl_db_code}')
    ret = pd.concat(df_array, axis=1)
    return ret


# In[99]:


# Here we are going to figure out what zip codes are in the specified county
# We will also use the logitude / latitude data for plotting

ZIP_CODE_DTYPE = {'zip_code': str}
ZIP_CODES = pd.read_csv('zip_codes_states.csv', dtype=ZIP_CODE_DTYPE)
county_df = ZIP_CODES['county'] == county
state_df = ZIP_CODES['state'] == state
data = ZIP_CODES[county_df & state_df]


# In[100]:


data.index = data['zip_code']


# In[101]:


# We will call our function to fetch the data from quandl
# There will be errors for data we cannot retrieve

real_estate_data = get_city_data(data['zip_code'], indicator)


# In[102]:


# Here we are just doing some transformation and massaging of the data we get back
# If you want more information please run the notebook on your own

transformed_real_estate_data = real_estate_data.transpose()
column_name = f'{indicator} {end_date}'


# In[103]:


transformed_real_estate_data[column_name] = transformed_real_estate_data[end_date].astype('float')


# In[104]:


concat_data = pd.concat([data, transformed_real_estate_data[column_name]], axis=1, sort=True)


# In[105]:


concat_data.dropna(inplace=True)


# In[106]:


concat_data['text'] = concat_data['zip_code'] + ' ' + concat_data[column_name].astype(str)


# In[107]:


df = concat_data


# In[108]:


# Here is our final result dataframe
# We can see for each zip code we have the longitude and latitude data
# As well as the Zillow Indicator and text label 

df.head()


# In[109]:


# Now it's just a matter of plotting using the Plotly Mapbox graph

plot_data = [
    go.Scattermapbox(
        lon = df['longitude'],
        lat = df['latitude'],
        mode='markers',
        marker=dict(
            reversescale = True,
            autocolorscale = False,
            size=17,
            opacity=0.85,
            colorscale = 'YlOrRd',
            cmin = df[column_name].min(),
            color = df[column_name],
            cmax = df[column_name].max(),
            colorbar=dict(
                title="Price"
            )
        ),
        text = df['text'],
        hoverinfo= 'text'
    )]

layout = go.Layout(
    title= f'{indicator} in {county}, {state} {end_date}',
    autosize=True,
    hovermode='closest',
    showlegend=False,
    mapbox=dict(
        accesstoken=mapbox_access_token,
        bearing=0,
        center=dict(
            lat=df.iloc[0]['latitude'],
            lon=df.iloc[0]['longitude']
        ),
        pitch=0,
        zoom=8,
        style='light'
    ),
)


# In[110]:


# Finally calling our plot function after declaring the layout

fig = dict( data=plot_data, layout=layout )
plotly.iplot( fig, filename=f'{indicator} {county} {state} {end_date}')


# In[111]:


# Output to div for blog post
# plotly.plot(fig, filename=f'{indicator} {county} {state} {end_date}', include_plotlyjs=False, output_type='div')


# In[ ]: