Regresión

In [22]:

    

import numpy as np #libreria de datos numericos

import matplotlib.pyplot as plt

import seaborn as sns
sns.set()

%matplotlib inline

import pandas as pd


    

In [36]:

    

a = 34
B = 100

x = np.random.normal(3,4,100)
y = a + B*x + np.random.normal(0,90,len(x)) # realizar correlaciones (error, varianza, len(x))


DF = pd.DataFrame({'x':x,'y':y})

DF.head()


    

    
Out[36]:






  

    

      

      
x
      
y
    
  
  

    

      
0
      
4.201274
      
523.079677
    
    

      
1
      
8.136716
      
732.464634
    
    

      
2
      
1.693414
      
370.752457
    
    

      
3
      
5.073713
      
630.349290
    
    

      
4
      
5.725034
      
666.353722
    
  

In [37]:

    

#plt.scatter(x, y)

sns.lmplot('x', 'y', DF)


    

    
Out[37]:





<seaborn.axisgrid.FacetGrid at 0x11c3c2150>






    

In [ ]:

    

Correlacion entre dos variables: y = c + Bx + e 

Tarea:
    Buscar datasets 
    
    
    
    statmodels # https://github.com/statsmodels/statsmodels A good repository with examples!
    Funciones
        .ols (ordinary least squares) # http://statsmodels.sourceforge.net/0.6.0/examples/notebooks/generated/ols.html
        .summary #http://statsmodels.sourceforge.net/0.6.0/generated/statsmodels.iolib.summary.Summary.html


    

In [4]: