anjulcodewiz / risk-modeling-using-sas Goto Github PK

Here is a step-by-step explanation of the SAS code:

Step 1: The code starts by turning on the ods graphics option and specifying the location of the librisk library.

Step 2: The proc import procedure is used to import a dataset named Credit_Data.csv from a specified location and create a new dataset named loan_data in the librisk library.

Step 3: The data step is used to manipulate the data and create new variables such as Delinquency, fico_avg, last_fico_avg, and income_category.

Step 4: The proc contents procedure is used to generate a report on the contents of the risk_data dataset.

Step 5: The proc means procedure is used to generate descriptive statistics for the numeric variables in the risk_data dataset.

Step 6: The proc freq procedure is used to generate frequency tables for the character variables in the risk_data dataset.

Step 7: The proc sgplot procedure is used to generate box plots and bar plots to visualize the relationship between different variables and the target variable (Delinquency).

Step 8: The code then removes outliers from the annual_inc variable and generates a histogram to visualize its distribution.

Step 9: The code also generates a donut plot to visualize the distribution of the purpose variable.

Step 10: The code uses the proc surveyselect procedure to generate a random sample of 10000 observations from the income_no_out dataset.

Step 11: The proc sgscatter procedure is used to generate a scatter plot to visualize the relationship between two numeric variables (EMI and Loan_Amount) and the target variable (Delinquency).

Step 12: The code then performs statistical data analysis using procedures such as proc corr, proc glm, and proc univariate.

Step 13: The code creates a new dataset named loan_data that contains only the relevant variables for further analysis.

Step 14: The code uses the proc stdize procedure to standardize some of the numeric variables in the loan_data dataset.

Step 15: The code then removes observations with missing values from the loan_data dataset.

Step 16: The code uses the proc reg procedure to fit a linear regression model to predict the target variable (Delinquency) using a set of predictor variables.

Step 17: The code uses the proc logistic procedure to fit a logistic regression model to predict the target variable (Delinquency) using a set of predictor variables, perform stepwise selection, and generate a ROC curve.

I hope this format makes it easier for you to understand how this SAS code works!