Fixing data imputation

Author: lucasthim

01_exploratory_analysis_pre_cleaning.ipynb

@@ -943,7 +943,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "- Most of the clients who did not meet the bank guidelines (Credit_History) had their loan applications rejected.\n",
+    "- Most of the clients who did not meet the bank guidelines for Credit_History had their loan applications rejected.\n",
     "- Most of the applicants income are under 5700.\n",
     "- Most of the co-applicants income are below 2200 and almost half of them have 0 income.\n",
     "- Most of the loan terms are 360 months.\n",

02_pre_processing.ipynb

@@ -13,15 +13,14 @@
    "source": [
     "Based on the conclusion of the Exploratory Data Analysis, we fill in some missing data assuming the following hypothesis:\n",
     "\n",
-    "- If a client has no information about the Credit_History and had their application rejected, they do not meet the bank guidelines (Credit_History = 0).\n",
-    "- Fill in the Self_employed with 'Employed', which is the most common case.\n",
+    "- Fill in the Credit_History with 1. \n",
     "- Fill in the LoanAmount_Terms with 360 for clients who have this value missing.\n",
     "- Fill in the CoapplicantIncome with 0."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 127,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -41,7 +40,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 128,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -50,7 +49,7 @@
        "(614, 14)"
       ]
      },
-     "execution_count": 128,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -70,15 +69,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 129,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [],
    "source": [
     "df_fill = df_import.copy()\n",
-    "# df_fill.loc[df_fill['Credit_History'].isnull(),'Credit_History'] = 1\n",
-    "df_fill.loc[(df_fill['Loan_Status'] == 'N') & (df_fill['Credit_History'].isnull()),'Credit_History'] = 0\n",
+    "df_fill.loc[df_fill['Credit_History'].isnull(),'Credit_History'] = 1\n",
     "df_fill.loc[df_fill['Loan_Amount_Term'].isnull(),'Loan_Amount_Term'] = 360\n",
-    "df_fill.loc[df_fill['Self_Employed'].isnull(),'Self_Employed'] = 'No'\n",
+    "# df_fill.loc[df_fill['Self_Employed'].isnull(),'Self_Employed'] = 'No'\n",
     "df_fill.loc[df_fill['CoapplicantIncome'].isnull(),'CoapplicantIncome'] = 0"
    ]
   },
@@ -91,7 +89,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 130,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -113,30 +111,30 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 131,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
-       "Gender                   530\n",
-       "Married                  530\n",
-       "Dependents               530\n",
-       "Education                530\n",
-       "Self_Employed            530\n",
-       "ApplicantIncome          530\n",
-       "CoapplicantIncome        530\n",
-       "LoanAmount               530\n",
-       "Loan_Amount_Term         530\n",
-       "Credit_History           530\n",
-       "Property_Area            530\n",
-       "Loan_Status              530\n",
-       "Base_Loan_Installment    530\n",
-       "Remaining_Income         530\n",
+       "Gender                   535\n",
+       "Married                  535\n",
+       "Dependents               535\n",
+       "Education                535\n",
+       "Self_Employed            535\n",
+       "ApplicantIncome          535\n",
+       "CoapplicantIncome        535\n",
+       "LoanAmount               535\n",
+       "Loan_Amount_Term         535\n",
+       "Credit_History           535\n",
+       "Property_Area            535\n",
+       "Loan_Status              535\n",
+       "Base_Loan_Installment    535\n",
+       "Remaining_Income         535\n",
        "dtype: int64"
       ]
      },
-     "execution_count": 131,
+     "execution_count": 7,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -155,7 +153,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 132,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -172,7 +170,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 133,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [
     {
@@ -204,7 +202,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 134,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -213,7 +211,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 135,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [
     {
@@ -466,7 +464,7 @@
        "9            2.0          1.0             194.444444          0.950142  "
       ]
      },
-     "execution_count": 135,
+     "execution_count": 11,
      "metadata": {},
      "output_type": "execute_result"
     }