Project Three: Likelihood of Pregnancy based on Race and Age

The project is to predict if there is a significant likelihood of pregnancy using the race and age of a unmarried woman as factors based on US data from 2006-2015. The different races being measured are All Races, Asian/Pacific Islander, African American/Black, Hispanic, Non-hispanic White, and All whites. The age groups measured are 15-19, 20-24, 25-29, 30-34, 35-39, and 40-44.

In [34]:
import os
try:
    inputFunc = raw_input
except NameError:
    inputFunc = input

import pandas as pd
from pandas.tseries.holiday import USFederalHolidayCalendar
from pandas.tseries.offsets import CustomBusinessDay
import numpy as np
 
import seaborn as sns
from statsmodels.formula.api import ols

from sklearn import linear_model
from sklearn import metrics

from sklearn.linear_model import LogisticRegression
from patsy import dmatrices

import matplotlib.pyplot as plt
%matplotlib inline
import numpy as np
from sklearn.metrics import accuracy_score
from sklearn.metrics import confusion_matrix

import matplotlib.pyplot as plt

import random


def evaluate(pred, labels_test):
    acc = accuracy_score(pred, labels_test)
    print ("Accuracey: %s"%acc)
    tn, fp, fn, tp = confusion_matrix(labels_test, pred).ravel()

    recall = tp / (tp + fp)
    percision = tp / (tp + fn)
    f1 = (2 / ((1/recall)+(1/percision)))

    print ("")
    print ("True Negatives: %s"%tn)
    print ("False Positives: %s"%fp)
    print ("False Negatives: %s"%fn)
    print ("True Positives: %s"%tp)
    print ("Recall: %s"%recall)
    print ("Precision: %s"%percision)
    print ("F1 Score: %s"%f1)

def plot_bound(Z_val,data,col1,col2,binary):

    x_min = float(data.iloc[:,[col1]].min())-float(data.iloc[:,[col1]].min())*0.10 
    x_max = float(data.iloc[:,[col1]].max()+float(data.iloc[:,[col1]].min())*0.10)
    y_min = 0.0; 
    y_max = float(training.iloc[:,[col2]].max())+float(training.iloc[:,[col2]].max())*0.10
    h_x = (x_max-x_min)/100  # step size in the mesh
    h_y = (y_max-y_min)/100  # step size in the mesh
    xx, yy = np.meshgrid(np.arange(x_min, x_max, h_x), np.arange(y_min, y_max, h_y))
    if binary == 1:
        Z = clf.predict(np.c_[xx.ravel(), yy.ravel()])   
        Z = np.where(Z=="Y",1,0)
    else:
        Z = clf.predict_proba(np.c_[xx.ravel(), yy.ravel()])[:, 1]

    Z = Z.reshape(xx.shape)
    plt.xlim(xx.min(), xx.max())
    plt.ylim(yy.min(), yy.max())
    plt.pcolormesh(xx, yy, Z)
    plt.show()

Data Cleaning

I retrieved this data from data.gov. Here I'm loading an edited version of unmarried_woman.csv from: https://catalog.data.gov/dataset/nonmarital-birth-rates-by-race-and-hispanic-origin-for-women-aged-15-44-united-states-1970. I used that data to create a table that relies on two variables to possibly affect the target variable.

In [35]:
raw_data_df = pd.read_csv('unmarried_woman.csv', parse_dates=[0]) 
raw_data_df.head()
Out[35]:
Year Age Race Birth_Rate
0 2015-01-01 15-19 years All Races 20.2
1 2015-01-01 15-19 years Asian or Pacific Islander total 5.5
2 2015-01-01 15-19 years Black total 31.5
3 2015-01-01 15-19 years Hispanic 31.6
4 2015-01-01 15-19 years Non-Hispanic white 13.9

Filtering Data and Training Data

I started filtering through the data, making sense of the different variables and the certain number of entries. I edited the entries for the variables so that they would all appear as numerical values. I also created a new column, which I named Pregnancy, which would gauge if there is a likelihood that the woman would become pregnant. This column uses the Birth_Rate column, and if that number is greater than 50, it would be "Y" for pregnancy likelihood. Otherwise it would be "N".

I made a table to train the data to use the variables age and race to gauge birth rate. This table falls under pregnant_class.df and will be used to analyze the data to make predictions.

In [36]:
print(raw_data_df["Age"].unique())
print(raw_data_df["Race"].unique())
print(raw_data_df["Birth_Rate"].unique())

['15-19 years' '20-24 years' '25-29 years' '30-34 years' '35-39 years'
 '40-44 years']
['All Races' 'Asian or Pacific Islander total' 'Black total' 'Hispanic'
 'Non-Hispanic white' 'White total']
[  20.2    5.5   31.5   31.6   13.9   18.8   22.     6.2   34.4   15.
   20.3   24.     7.2   38.5   37.6   16.2   21.9   26.7    8.1   43.4
   41.8   17.8   24.1   28.4    8.6   46.7   44.7   25.5   31.1    9.2
   50.8   50.    27.9   34.    10.6   55.9   56.7   21.8   30.4   35.9
   11.4   59.7   62.4   22.5   31.9   36.5   11.9   61.3   65.4   22.6
   32.3   35.5   12.    61.1   65.9   21.6   17.9   94.2   87.9   43.3
   54.6   61.6   19.4   97.4   92.3   44.4   63.1   20.7  100.7   93.1
   45.6   56.9   64.7  103.5   96.5   46.6   58.3   66.7   22.3  106.9
  100.6   47.8   60.1   70.    23.2  112.6  110.5   49.5   63.4   74.4
   25.3  119.5  125.4   51.4   67.3   78.1   26.4  124.   141.    52.9
   70.9   79.8   27.1  125.3  153.8   53.4   72.6   79.1   26.8  125.5
  155.1   52.6   71.7   66.9   28.8   91.9  109.1   48.8   63.3   67.6
   31.4   93.3  111.2   49.1   63.5  112.    47.6   62.5   67.2   35.2
   91.2  113.2   63.2   67.8   35.1   92.4  116.2   63.8   69.2   35.
   92.5  123.9   48.    65.8   73.    36.2   95.6  139.4   49.3   69.7
   75.7   33.5   97.   151.1   73.1   76.9   99.1  161.1   50.7   74.3
   75.4   30.5   98.   160.7   72.4   60.3   47.3   63.9  101.4   59.8
   58.1   47.    67.    97.3   43.1   57.7   56.6   44.1   60.6  100.1
   40.9   56.2   56.3   43.6   59.6  103.9   40.2   42.3   59.1  106.
   39.2   40.    58.6  105.8   38.7   56.8   57.1  112.4   37.9   58.8
  121.6   38.4   60.    58.    36.9   59.9  127.    37.2   59.5   55.3
   58.4  122.9   55.5   34.1   33.6   32.6   24.8   34.8   33.4   36.    32.
   58.2   24.2   31.8   30.9   30.1   29.7   57.6   29.9   27.    28.9
   30.6   29.6   26.6   27.8   61.7   30.7   27.3   63.7   20.    30.2
   30.3   28.5   20.4   28.7   28.    64.9   19.2   29.1    9.    11.2
    9.1   18.1    6.8    8.2    8.5   12.6    8.8   16.8    6.5    7.6
    8.3    8.4    6.1    7.7   12.1   16.5    9.9    7.8    6.     8.
   16.3    5.8   17.1    7.5   10.8    5.2    7.3    7.4   14.9    4.6
    9.4   14.8    4.5    6.3]
In [37]:
raw_data_df[raw_data_df["Age"]=='20-24 years'].head()
Out[37]:
Year Age Race Birth_Rate
60 2015-01-01 20-24 years All Races 59.7
61 2015-01-01 20-24 years Asian or Pacific Islander total 17.9
62 2015-01-01 20-24 years Black total 94.2
63 2015-01-01 20-24 years Hispanic 87.9
64 2015-01-01 20-24 years Non-Hispanic white 43.3
In [38]:
print("Size of entire table: %s "%len(raw_data_df))
print("Size of entires matching filter: %s "%len(raw_data_df[raw_data_df["Age"]=="20-24 years"]))

Size of entire table: 360 
Size of entires matching filter: 60
In [39]:
raw_data_df.loc[raw_data_df['Age'] == '15-19 years', 'Age'] = 18
raw_data_df.loc[raw_data_df['Age'] == '20-24 years', 'Age'] = 21
raw_data_df.loc[raw_data_df['Age'] == '25-29 years', 'Age'] = 26
raw_data_df.loc[raw_data_df['Age'] == '30-34 years', 'Age'] = 30
raw_data_df.loc[raw_data_df['Age'] == '35-39 years', 'Age'] = 35
raw_data_df.loc[raw_data_df['Age'] == '40-44 years', 'Age'] = 42
raw_data_df.loc[raw_data_df['Race'] == 'All Races', 'Race'] = 1
raw_data_df.loc[raw_data_df['Race'] == 'Asian or Pacific Islander total', 'Race'] = 2
raw_data_df.loc[raw_data_df['Race'] == 'Black total', 'Race'] = 3
raw_data_df.loc[raw_data_df['Race'] == 'Hispanic', 'Race'] = 4
raw_data_df.loc[raw_data_df['Race'] == 'Non-Hispanic white', 'Race'] = 5
raw_data_df.loc[raw_data_df['Race'] == 'White total', 'Race'] = 6
raw_data_df.head()
Out[39]:
Year Age Race Birth_Rate
0 2015-01-01 18 1 20.2
1 2015-01-01 18 2 5.5
2 2015-01-01 18 3 31.5
3 2015-01-01 18 4 31.6
4 2015-01-01 18 5 13.9
In [40]:
raw_data_df["Pregnant"] = "N"
raw_data_df.head()
Out[40]:
Year Age Race Birth_Rate Pregnant
0 2015-01-01 18 1 20.2 N
1 2015-01-01 18 2 5.5 N
2 2015-01-01 18 3 31.5 N
3 2015-01-01 18 4 31.6 N
4 2015-01-01 18 5 13.9 N
In [46]:
raw_data_df.loc[raw_data_df['Birth_Rate'] >= 60, 'Pregnant'] = "Y"
raw_data_df.head()
Out[46]:
Year Age Race Birth_Rate Pregnant
0 2015-01-01 18 1 20.2 N
1 2015-01-01 18 2 5.5 N
2 2015-01-01 18 3 31.5 N
3 2015-01-01 18 4 31.6 N
4 2015-01-01 18 5 13.9 N
In [47]:
# I'm now going to make a table to be used in training some models
In [48]:
# The set will be for classifiers where the target is a class.

pregnant_class_df = raw_data_df[[
                               'Pregnant', 
                               'Age',
                                'Race'
                               ]].copy()
pregnant_class_df.head()
Out[48]:
Pregnant Age Race
0 N 18 1
1 N 18 2
2 N 18 3
3 N 18 4
4 N 18 5

Training and Validation

Above I created the dataset worth exploring:

  • pregnancy_class_df. The data needed to access pregnancy likelihood as a categorical variable.

pregnant_class_df

In [49]:
data = pregnant_class_df
holdout = data.sample(frac=0.20)
training = data.loc[~data.index.isin(holdout.index)]

# Define the target (y) and feature(s) (X)
features_train = training.drop("Pregnant", axis=1).as_matrix(columns=None)
labels_train = training["Pregnant"].as_matrix(columns=None)

features_test = holdout.drop( "Pregnant", axis=1).as_matrix(columns=None)
labels_test = holdout["Pregnant"].as_matrix(columns=None)

# What percentage of the time is target Y?
print("Percentage of Ys: %s\n"%(len(data[data["Pregnant"]=="Y"])/len(data)))

#### initial visualization
feature_1_no = [features_test[ii][0] for ii in range(0, len(features_test)) if labels_test[ii]=="N"]
feature_2_no = [features_test[ii][1] for ii in range(0, len(features_test)) if labels_test[ii]=="N"]
feature_1_yes = [features_test[ii][0] for ii in range(0, len(features_test)) if labels_test[ii]=="Y"]
feature_2_yes = [features_test[ii][1] for ii in range(0, len(features_test)) if labels_test[ii]=="Y"]
plt.scatter(feature_1_yes, feature_2_yes, color = "g", label="Likely Pregnant")
plt.scatter(feature_1_no, feature_2_no, color = "r", label="Unlikely Pregnant")
plt.legend()
plt.xlabel("Age")
plt.ylabel("Race")
plt.show()

Percentage of Ys: 0.2861111111111111
In [50]:
# Logistic Regression
model = LogisticRegression(fit_intercept = False, C = 1e9)
clf = model.fit(features_train, labels_train)
pred = clf.predict(features_test)
print("Logistic Regression")
evaluate(pred, labels_test)  
plot_bound("Y",holdout,1,2,0)


# Test some spot
x_test = 70
y_test = 16000
print("")
print(clf.predict([[x_test,y_test]])[0])
print(clf.predict_proba([[x_test,y_test]])[0][1])
print("")

from sklearn import tree
clf = tree.DecisionTreeClassifier(min_samples_split=40)
clf = clf.fit(features_train, labels_train)
pred = clf.predict(features_test)
print("\nDecision Tree")
evaluate(pred, labels_test)
plot_bound("Y",holdout,1,2,0)



from sklearn.ensemble import RandomForestClassifier
clf = RandomForestClassifier()
clf = clf.fit(features_train, labels_train)
pred = clf.predict(features_test)
print("Random Forest")
evaluate(pred, labels_test)  
plot_bound("Y",holdout,1,2,0)


from sklearn.svm import SVC
clf = SVC(kernel="rbf",probability=True)
clf = clf.fit(features_train, labels_train)
pred = clf.predict(features_test)
print("SVM")
evaluate(pred, labels_test)  

Logistic Regression
Accuracey: 0.75

True Negatives: 54
False Positives: 0
False Negatives: 18
True Positives: 0
Recall: nan
Precision: 0.0
F1 Score: nan

C:\Users\Loren\Anaconda3\lib\site-packages\ipykernel_launcher.py:37: RuntimeWarning: invalid value encountered in long_scalars
C:\Users\Loren\Anaconda3\lib\site-packages\ipykernel_launcher.py:39: RuntimeWarning: divide by zero encountered in double_scalars

Y
1.0


Decision Tree
Accuracey: 0.819444444444

True Negatives: 54
False Positives: 0
False Negatives: 13
True Positives: 5
Recall: 1.0
Precision: 0.277777777778
F1 Score: 0.434782608696

Random Forest
Accuracey: 0.902777777778

True Negatives: 52
False Positives: 2
False Negatives: 5
True Positives: 13
Recall: 0.866666666667
Precision: 0.722222222222
F1 Score: 0.787878787879

SVM
Accuracey: 0.902777777778

True Negatives: 50
False Positives: 4
False Negatives: 3
True Positives: 15
Recall: 0.789473684211
Precision: 0.833333333333
F1 Score: 0.810810810811

Result analysis

Based on the result, the analysis is that there is a possibility at most ages and races that pregnancy will occur. However, the most common age was the 25-29 range, and the most common races to get pregnant were Hispanics and African-Americans. The logisitical regression was the least accurate with many false negatives, meaning that it was predicting that the pregnancy probability would be less than was suggested in the data. There was the fact that there were several variables being tested and spewing out multiple results for each data point, which may have resulted in the inconsistencies. There was also fewer false positives, proving that the data was skewed toward the side of less likely. Overall, the conclusion that all the graphs and analysis pointed to was that unmarried Hispanic and African-American women were more likely to become pregnant.