
%pylab inline
import pandas as pd

Populating the interactive namespace from numpy and matplotlib


#Import the SBA Loans dataset

sba = pd.read_csv("DSTMAA_data/SBA.csv")
print(sba.columns)
print(sba.shape)
sba.head()

Index(['LoanID', 'GrossApproval', 'SBAGuaranteedApproval', 'subpgmdesc',
       'ApprovalFiscalYear', 'InitialInterestRate', 'TermInMonths',
       'ProjectState', 'BusinessType', 'LoanStatus', 'RevolverStatus',
       'JobsSupported'],
      dtype='object')
(527700, 12)


#Create the dependent variable 
Y = sba.SBAGuaranteedApproval.astype("float")/sba.GrossApproval.astype("float")
X = sba[['GrossApproval', 'ApprovalFiscalYear', 'InitialInterestRate', 'TermInMonths',
        'RevolverStatus','JobsSupported']]


x1 = pd.get_dummies(sba.subpgmdesc)
x1 = x1.drop(x1.columns[0],axis=1)
X = pd.concat([X,x1],axis=1)
X.head()


x1 = pd.get_dummies(sba.BusinessType)
x1 = x1.drop(x1.columns[0],axis=1)
X = pd.concat([X,x1],axis=1)
X.head()


#SAVE TO PICKLE
import pickle
SBAdata = {'sba': X}
pickle.dump(SBAdata, open( "DSTMAA_data/SBAdata.p", "wb" ))


#Run regression
from sklearn import linear_model
reg = linear_model.LinearRegression()
reg.fit(X,Y)
yhat = reg.predict(X)


from sklearn.metrics import mean_squared_error, r2_score
# The coefficients
print('Coefficients: \n', reg.coef_)
# The mean squared error
print("Mean squared error: %.2f" % mean_squared_error(Y,yhat))
# Explained variance score: 1 is perfect prediction
print('Variance score: %.2f' % r2_score(Y,yhat))

Coefficients: 
 [-1.57772017e-08 -2.93254437e-03 -3.36804150e-03 -5.01226990e-05
 -4.24578025e-04 -2.01392544e-05  9.66450791e-02  1.18528756e-01
  4.73548794e-02 -1.64573288e-02 -2.36558347e-01  6.51446091e-02
  1.21989162e-01  1.85657957e-01  6.89418169e-02  1.02387596e-01
  1.72723592e-01  1.02752017e-01  4.41276709e-02  1.05984173e-01
  5.08275851e-02  3.04713903e-02  2.42619720e-03 -6.42487723e-04]
Mean squared error: 0.00
Variance score: 0.92


import statsmodels.api as sm
import statsmodels.formula.api as smf

X = sm.add_constant(X, prepend=True)
results = smf.OLS(Y,X).fit()

print(results.summary())

/home/srdas/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py:52: FutureWarning: Method .ptp is deprecated and will be removed in a future version. Use numpy.ptp instead.
  return getattr(obj, method)(*args, **kwds)

                            OLS Regression Results                            
==============================================================================
Dep. Variable:                      y   R-squared:                       0.921
Model:                            OLS   Adj. R-squared:                  0.921
Method:                 Least Squares   F-statistic:                 2.575e+05
Date:                Wed, 24 Apr 2019   Prob (F-statistic):               0.00
Time:                        13:46:37   Log-Likelihood:             8.9683e+05
No. Observations:              527700   AIC:                        -1.794e+06
Df Residuals:                  527675   BIC:                        -1.793e+06
Df Model:                          24                                         
Covariance Type:            nonrobust                                         
=====================================================================================================================
                                                        coef    std err          t      P>|t|      [0.025      0.975]
---------------------------------------------------------------------------------------------------------------------
const                                                 6.6665      0.058    115.637      0.000       6.554       6.780
GrossApproval                                     -1.578e-08   1.54e-10   -102.493      0.000   -1.61e-08   -1.55e-08
ApprovalFiscalYear                                   -0.0029   2.86e-05   -102.639      0.000      -0.003      -0.003
InitialInterestRate                                  -0.0034   3.32e-05   -101.516      0.000      -0.003      -0.003
TermInMonths                                      -5.012e-05   1.14e-06    -43.935      0.000   -5.24e-05   -4.79e-05
RevolverStatus                                       -0.0004      0.000     -2.610      0.009      -0.001      -0.000
JobsSupported                                     -2.014e-05   1.57e-06    -12.865      0.000   -2.32e-05   -1.71e-05
Community Advantage Initiative                        0.0966      0.004     24.790      0.000       0.089       0.104
Community Express                                     0.1185      0.004     31.434      0.000       0.111       0.126
Contract Guaranty                                     0.0474      0.004     10.792      0.000       0.039       0.056
EXPORT IMPORT HARMONIZATION                          -0.0165      0.011     -1.448      0.148      -0.039       0.006
FA$TRK (Small Loan Express)                          -0.2366      0.004    -62.960      0.000      -0.244      -0.229
Guaranty                                              0.0651      0.004     17.326      0.000       0.058       0.073
Gulf Opportunity                                      0.1220      0.004     31.621      0.000       0.114       0.130
International Trade - Sec, 7(a) (16)                  0.1857      0.004     45.123      0.000       0.178       0.194
Lender Advantage Initiative                           0.0689      0.004     18.275      0.000       0.062       0.076
Patriot Express                                       0.1024      0.004     27.063      0.000       0.095       0.110
Revolving Line of Credit Exports - Sec. 7(a) (14)     0.1727      0.004     43.418      0.000       0.165       0.181
Rural Lender Advantage                                0.1028      0.004     26.899      0.000       0.095       0.110
Seasonal Line of Credit                               0.0441      0.005      8.256      0.000       0.034       0.055
Small Asset Based                                     0.1060      0.007     15.688      0.000       0.093       0.119
Small General Contractors - Sec. 7(a) (9)             0.0508      0.005      9.417      0.000       0.040       0.061
Standard Asset Based                                  0.0305      0.004      7.862      0.000       0.023       0.038
INDIVIDUAL                                            0.0024      0.000     14.331      0.000       0.002       0.003
PARTNERSHIP                                          -0.0006      0.000     -1.456      0.145      -0.002       0.000
==============================================================================
Omnibus:                   239343.300   Durbin-Watson:                   1.555
Prob(Omnibus):                  0.000   Jarque-Bera (JB):         13407205.167
Skew:                           1.403   Prob(JB):                         0.00
Kurtosis:                      27.533   Cond. No.                     5.52e+08
==============================================================================

Warnings:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The condition number is large, 5.52e+08. This might indicate that there are
strong multicollinearity or other numerical problems.

	LoanID	GrossApproval	SBAGuaranteedApproval	subpgmdesc	ApprovalFiscalYear	InitialInterestRate	TermInMonths	ProjectState	BusinessType	LoanStatus	RevolverStatus	JobsSupported
0	733784	50000	25000	FA$TRK (Small Loan Express)	2006	11.25	84	IN	CORPORATION	CANCLD	1	4
1	733785	35000	17500	FA$TRK (Small Loan Express)	2006	12.00	84	IL	CORPORATION	CANCLD	0	3
2	733786	15000	7500	FA$TRK (Small Loan Express)	2006	12.00	84	WV	INDIVIDUAL	CANCLD	0	4
3	733787	16000	13600	Community Express	2006	11.50	84	MD	CORPORATION	PIF	0	1
4	733788	16000	13600	Community Express	2006	11.50	84	MD	CORPORATION	CANCLD	0	1

Linear Regression Models¶

Predict the guaranteed percentage as a function of the other variables¶

Run the regression using scikit-learn¶

Run the regression using statsmodels¶