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

# <a href="https://colab.research.google.com/github/cyrus723/my-first-binder/blob/main/simple_regression_1.ipynb" target="_parent"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open In Colab"/></a>

# Source of data: FactSets (All ETFs with some cleaning)

# In[182]:


import numpy as np
import pandas as pd
from pylab import mpl, plt
import statsmodels.formula.api as smf
import statsmodels.api as sm
plt.style.use('seaborn')
mpl.rcParams['font.family'] = 'serif'
get_ipython().run_line_magic('matplotlib', 'inline')


# In[2]:


import os 
os.getcwd()


# In[120]:


filename = 'https://raw.githubusercontent.com/cyrus723/my-first-binder/main/factsets_etf.csv'  


# In[121]:


data = pd.read_csv(filename)  


# In[122]:


data.info()  


# In[123]:


data['scaled_aum'] = np.log(data['aum'])


# In[124]:


data['aum_million'] = data['aum'] / 1000000


# In[125]:


data.head()  


# In[126]:


data.tail()  


# ### Summary Statistics

# In[127]:


data.info()  


# In[128]:


data.describe().round(2)  


# In[129]:


data.aggregate([min,  
                np.mean,  
                np.std,  
                np.median,  
                max]  
).round(2)


# In[131]:


data.info()


# In[132]:


data2 = data.groupby(by='class').mean()
data2


# In[138]:


data['class'] = data['class'].replace(['Asset Allocation','Fixed Income' ], ['Asset_Allocation', "Fixed_Income"])


# In[143]:


data2 = data.groupby(by='class').mean()
data2


# In[144]:


data3 = data.rename({'class': 'asset_class'}, axis=1, inplace=False)


# In[145]:


data.info()


# In[146]:


data3.info()


# In[149]:


data4 = pd.get_dummies(data3, columns=['asset_class'])


# In[150]:


data4.info()


# ## Regression Analysis

# In[157]:


data5 = data4[['exp', 
              'scaled_aum', 
              'beta', 
              'asset_class_Alternatives',
              'asset_class_Commodities',
              'asset_class_Currency',
              'asset_class_Equity',
              'asset_class_Fixed_Income']].dropna()


# In[158]:


data5.describe()


# In[159]:


data6 = data5.loc[data5['exp'] <= 3]


# In[160]:


data6.describe()


# In[ ]:





# In[161]:


data7 = data6[['exp', 'scaled_aum']].dropna()


# In[162]:


pd.plotting.scatter_matrix(data7,  
                           alpha=0.2,  
                           diagonal='hist',  
                           hist_kwds={'bins': 35},  
                           figsize=(10, 6));
# plt.savefig('../../images/ch08/fts_11.png');


# In[163]:


data7 = data6[['exp', 'beta']].dropna()


# In[164]:


pd.plotting.scatter_matrix(data7,  
                           alpha=0.2,  
                           diagonal='hist',  
                           hist_kwds={'bins': 35},  
                           figsize=(10, 6));
# plt.savefig('../../images/ch08/fts_11.png');


# In[ ]:





# ### OLS Regression

# In[165]:


reg = np.polyfit(data6['exp'], data6['scaled_aum'], deg=1)  


# In[166]:


ax = data6.plot(kind='scatter', x='scaled_aum', y='exp', figsize=(10, 6))  
ax.plot(data6['scaled_aum'], np.polyval(reg, data6['scaled_aum']), 'r', lw=2);  
# plt.savefig('../../images/ch08/fts_12.png');


# In[183]:


formula = 'exp ~ scaled_aum + beta + asset_class_Commodities+	asset_class_Currency+	asset_class_Equity	+asset_class_Fixed_Income'
results = smf.ols(formula, data6).fit()
print(results.summary())


# ### Correlation

# In[171]:


data2.corr()  


# In[179]:


X = data6[['scaled_aum', 'beta','asset_class_Commodities','asset_class_Currency','asset_class_Equity','asset_class_Fixed_Income']]
y = data6['exp']


# In[184]:


X_ols = sm.add_constant(X)
model = sm.OLS(y, X_ols).fit()
print(model.summary())


# In[ ]:





# In[185]:


beta = np.linalg.inv(X_ols.T.dot(X_ols)).dot(X_ols.T.dot(y))
pd.Series(beta, index=X_ols.columns)