result_prediction.py

# coding: utf-8

# # *CSIT Board exam percentage score predition based on data of 120 students Midterm, FinalTerm, Attendance, InternalMarking, AssignmentScores*

# In[ ]:

# some basic imports

# pandas for csv data reading
import pandas as pd

# We'll use tree model (DecisionTree)
from sklearn import tree

# To print the accuracy score
from sklearn.metrics import accuracy_score


# In[ ]:

# Reading training data
train_df = pd.read_csv("trainingData.csv")

# Reading testing data
test_df = pd.read_csv("testingData.csv")
train_df.head()


# In[ ]:

# the normalize function since our class has numerical data, and we need to convert to categorical data
def normalize(num):
    if num >= 0.80:
        return "distinction"
    elif num >= 0.60 and num < 0.80:
        return "first div"
    elif num >= 0.40 and num < 0.60:
        return "second div"
    else:
        return "fail"


# In[ ]:

# X_train has first 5 attributes and y_train has the class label
# X is 2D and y is 1D
X_train, y_train = train_df.iloc[:,:-1], train_df.iloc[:, -1]
X_test, y_test = test_df.iloc[:,:-1], test_df.iloc[:,-1]

# Applying normalize funxtion to out class label series data
y_train = y_train.apply(normalize)
y_test = y_test.apply(normalize)
print X_train.shape, y_train.shape, X_test.shape, y_test.shape


# In[ ]:

# define our tree model
model = tree.DecisionTreeClassifier()

# fit our traing data into tree mddel
model.fit(X_train, y_train)

# predict on our test data
predictions = model.predict(X_test)

# print accuracy score based on testing data and prediction data
print accuracy_score(y_test, predictions)