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

# # Correlation Matrix in Python Using Downloaded Stock Data
# #### Assumes you have python and a number of supporting libraries installed. The easiest way to get most everything you need is to install the anaconda scientific platform by by visiting http://anaconda.com
# #### You will also need to install fix_yahoo_finance separately in order to download data from Yahoo Finance.  From a command prompt enter pip install fix_yahoo_finance
# 
# 

# ### 1. Set up environment 

# In[24]:


import numpy as np
import pandas as pd
import fix_yahoo_finance as fyf


# #### Note fix_yahoo_finance is billed a a temporary solution to recent API changes at Yahoo Finance. Pandas does have a complimentary library, pandas_datareader that allows that downloading of data from a number of sources. See http://pandas-datareader.readthedocs.io/en/latest/ for more information.

# ### 2. Download data for FANG stocks

# In[25]:


stocks = 'FB AMZN NFLX GOOG GLD'
stocks = stocks.split()
data = fyf.download(stocks, '2017-06-01')['Close']
data.head()


# In[26]:


import matplotlib.pyplot as plt
get_ipython().run_line_magic('matplotlib', 'inline')
plt.plot(data)
plt.legend(['AMZN', 'FB', 'GLD', 'GOOG', 'NFLX'], loc=2)


# ### 3. Transform closing price into instantaneous rates of return

# In[28]:


returns = pd.DataFrame()
for stock in data:
    if stock not in returns:
        returns[stock] = np.log(data[stock]).diff()
returns = returns[1:]        
returns.head()        


# ### 4. Generate descriptives with pandas method [describe]

# In[29]:


returns.describe()


# ### 5. Create correlation matrix

# In[30]:


returns.corr()


# In[32]:


returns.corr()['AMZN'].sort_values(ascending=False)


# ### 6. Graphically display correlations

# In[33]:


from pandas.plotting import scatter_matrix
scatter_matrix(returns, figsize=(16,12), alpha=0.3)


# In[ ]: