Home

Model About Datasets

import numpy as np
import warnings
import random

warnings.filterwarnings("ignore")
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import plotly.graph_objects as go

# model
from sklearn.metrics import confusion_matrix
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.ensemble import RandomForestClassifier

# pipeline
from sklearn.pipeline import Pipeline, make_pipeline
from sklearn.model_selection import GridSearchCV

data = pd.read_excel("onlinefoods.xlsx")
data.head(10)

	Age	Gender	Marital Status	Occupation	Monthly Income	Educational Qualifications	Family size	latitude	longitude	Pin code	Output	Feedback
0	20	Female	Single	Student	No Income	Post Graduate	4	12.9766	77.5993	560001	Yes	Positive
1	24	Female	Single	Student	Below Rs.10000	Graduate	3	12.9770	77.5773	560009	Yes	Positive
2	22	Male	Single	Student	Below Rs.10000	Post Graduate	3	12.9551	77.6593	560017	Yes	Negative
3	22	Female	Single	Student	No Income	Graduate	6	12.9473	77.5616	560019	Yes	Positive
4	22	Male	Single	Student	Below Rs.10000	Post Graduate	4	12.9850	77.5533	560010	Yes	Positive
5	27	Female	Married	Employee	More than 50000	Post Graduate	2	12.9299	77.6848	560103	Yes	Positive
6	22	Male	Single	Student	No Income	Graduate	3	12.9770	77.5773	560009	Yes	Positive
7	24	Female	Single	Student	No Income	Post Graduate	3	12.9828	77.6131	560042	Yes	Positive
8	23	Female	Single	Student	No Income	Post Graduate	2	12.9766	77.5993	560001	Yes	Positive
9	23	Female	Single	Student	No Income	Post Graduate	4	12.9854	77.7081	560048	Yes	Positive

data.dtypes

Age                             int64
Gender                         object
Marital Status                 object
Occupation                     object
Monthly Income                 object
Educational Qualifications     object
Family size                     int64
latitude                      float64
longitude                     float64
Pin code                        int64
Output                         object
Feedback                       object
dtype: object

data.isnull().sum()

Age                           0
Gender                        0
Marital Status                0
Occupation                    0
Monthly Income                0
Educational Qualifications    0
Family size                   0
latitude                      0
longitude                     0
Pin code                      0
Output                        0
Feedback                      0
dtype: int64

plt.figure(figsize=(10, 8))
plt.title("Online Food Order Decisions Based on the Age of the Customer")
sns.countplot(x="Age", data=data, hue="Feedback", palette="Set3")

<Axes: title={'center': 'Online Food Order Decisions Based on the Age of the Customer'}, xlabel='Age', ylabel='count'>

plt.figure(figsize=(10, 8))
plt.title("Online Food Order Decisions Based on the Size of the Family")
sns.countplot(x="Family size", data=data, hue="Feedback", palette="Set3_r")

<Axes: title={'center': 'Online Food Order Decisions Based on the Size of the Family'}, xlabel='Family size', ylabel='count'>

plt.figure(figsize=(10, 8))
plt.title(
    "Online Food Order Decisions Based on the Educational Qualifications of the Customer"
)
sns.countplot(x="Educational Qualifications", data=data, hue="Feedback", palette="Set2")

<Axes: title={'center': 'Online Food Order Decisions Based on the Educational Qualifications of the Customer'}, xlabel='Educational Qualifications', ylabel='count'>

plt.figure(figsize=(10, 8))
plt.title(
    "Online Food Order Decisions Based on the Educational Qualifications by the Occupation of the Customer"
)
sns.countplot(
    x="Educational Qualifications", data=data, hue="Occupation", palette="Set2_r"
)

<Axes: title={'center': 'Online Food Order Decisions Based on the Educational Qualifications by the Occupation of the Customer'}, xlabel='Educational Qualifications', ylabel='count'>

female = len(data[data["Gender"] == "Female"])
male = len(data[data["Gender"] == "Male"])
print(male, female)
data_gender = [female, male]
labels = ["Female", "Male"]

colors = sns.color_palette("pastel")[0:5]  # seaborn color palette to use

plt.figure(figsize=(10, 8))
plt.title("Distribution of Customer's Gender")
plt.pie(data_gender, labels=labels, colors=colors, autopct="%.0f%%")
plt.show()

222 166

plt.figure(figsize=(10, 8))
plt.title("Educational Qualifications Based on the Gender of the Customer")
sns.countplot(x="Occupation", data=data, hue="Gender", palette="Pastel2")

<Axes: title={'center': 'Educational Qualifications Based on the Gender of the Customer'}, xlabel='Occupation', ylabel='count'>

import seaborn as sns
import plotly.graph_objects as go

reorder = data[data["Feedback"] == "Positive"]

gender_data = reorder["Gender"].value_counts()

label = gender_data.index

counts = gender_data.values

print(label, counts)

colors = sns.color_palette("pastel")[2:4]


fig = go.Figure(data=[go.Pie(labels=label, values=counts)])

fig.update_layout(title_text="Which Gender is More Likely to Order Online Again?")

fig.update_traces(

    hoverinfo="label+percent", textinfo="value", marker=dict(colors=colors)
)


fig.show()

Index(['Male', 'Female'], dtype='object', name='Gender') [178 139]

plt.figure(figsize=(10, 8))
plt.title("Educational Qualifications Based on the Gender of the Customer")
sns.countplot(x="Monthly Income", data=data, hue="Feedback", palette="Pastel2_r")

<Axes: title={'center': 'Educational Qualifications Based on the Gender of the Customer'}, xlabel='Monthly Income', ylabel='count'>

reorder = data[data["Feedback"] == "Positive"]
status_data = reorder["Marital Status"].value_counts()
label = status_data.index
counts = status_data.values

fig = go.Figure(data=[go.Pie(labels=label, values=counts)])
fig.update_layout(title_text="What is the Marital Status of Customers?")
fig.update_traces(
    hoverinfo="label+percent", textinfo="value", marker=dict(colors=colors)
)

fig.show()

data["Monthly Income"].unique()

array(['No Income', 'Below Rs.10000', 'More than 50000', '10001 to 25000',
       '25001 to 50000'], dtype=object)

data["Gender"] = data["Gender"].map({"Male": 0, "Female": 1})  # male or female

data["Marital Status"] = data["Marital Status"].map(
    {"Married": 0, "Single": 1, "Prefer not to say": 2}
)


data["Occupation"] = data["Occupation"].replace(
    to_replace=["Employee", "Self Employeed"], value=1
)  # employed
data["Occupation"] = data["Occupation"].replace(
    to_replace=["Student", "House wife"], value=0
)  # unemployed


data["Educational Qualifications"] = data["Educational Qualifications"].map(
    {"Graduate": 1, "Post Graduate": 2, "Ph.D": 3, "School": 4, "Uneducated": 5}
)


data["Monthly Income"] = data["Monthly Income"].replace(
    to_replace=["No Income"], value=0
)  # no income
data["Monthly Income"] = data["Monthly Income"].replace(
    to_replace=[
        "Below Rs.10000",
        "More than 50000",
        "25001 to 50000",
        "10001 to 25000",
    ],
    value=1,
)  # has an income


data["Output"] = data["Output"].map({"No": 0, "Yes": 1})  # no or yes

data["Feedback"] = data["Feedback"].map(
    {"Negative ": 0, "Positive": 1}
)  # negative or positive

data.head()

	Age	Gender	Marital Status	Occupation	Monthly Income	Educational Qualifications	Family size	latitude	longitude	Pin code	Output	Feedback
0	20	1	1	0	0	2	4	12.9766	77.5993	560001	1	1
1	24	1	1	0	1	1	3	12.9770	77.5773	560009	1	1
2	22	0	1	0	1	2	3	12.9551	77.6593	560017	1	0
3	22	1	1	0	0	1	6	12.9473	77.5616	560019	1	1
4	22	0	1	0	1	2	4	12.9850	77.5533	560010	1	1

X = data.drop("Feedback", axis=1)
y = data["Feedback"]

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=101
)

print("Shape of train dataset : ", X_train.shape)
print("Shape of test dataset : ", X_test.shape)

Shape of train dataset :  (310, 11)
Shape of test dataset :  (78, 11)

rfc = RandomForestClassifier(n_estimators=100)

rfc.fit(X_train, y_train)

print(rfc.score(X_test, y_test))

0.7948717948717948

pred = rfc.predict(X_test)

print(confusion_matrix(pred, y_test))

plt.figure(figsize=(10, 8))
sns.heatmap(confusion_matrix(pred, y_test), annot=True)

[[ 5  5]
 [11 57]]

<Axes: >

# Evaluating a score by cross-validation
# cv determines the cross-validation splitting strategy
scores = cross_val_score(rfc, X_train, y_train, cv=5)

# average score
print("Accuracy: ", scores.mean(), scores.std() * 2)

Accuracy:  0.867741935483871 0.06255070783762999

parameters = {"randomforestclassifier__n_estimators": (20, 50, 100)}

pipeline = make_pipeline(RandomForestClassifier())

pipeline

Pipeline(steps=[('randomforestclassifier', RandomForestClassifier())])

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

gridsearch = GridSearchCV(pipeline, parameters, verbose=1, n_jobs=-1)

gridsearch.fit(X_train, y_train)

Fitting 5 folds for each of 3 candidates, totalling 15 fits

GridSearchCV(estimator=Pipeline(steps=[('randomforestclassifier',
                                        RandomForestClassifier())]),
             n_jobs=-1,
             param_grid={'randomforestclassifier__n_estimators': (20, 50, 100)},
             verbose=1)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

print("Best score %0.3f" % gridsearch.best_score_)
print("Best parameters set: ")
best_parameters = gridsearch.best_estimator_.get_params()

for params in sorted(parameters.keys()):
    print("\t%s: %r" % (params, best_parameters[params]))

Best score 0.877
Best parameters set: 
	randomforestclassifier__n_estimators: 100

r = random.randint(0, len(data))
print(r)
new_customer = data.drop("Feedback", axis=1).iloc[r]
new_customer

4

Age                               22.0000
Gender                             0.0000
Marital Status                     1.0000
Occupation                         0.0000
Monthly Income                     1.0000
Educational Qualifications         2.0000
Family size                        4.0000
latitude                          12.9850
longitude                         77.5533
Pin code                      560010.0000
Output                             1.0000
Name: 4, dtype: float64

rfc.predict(new_customer.values.reshape(1, -1))  # predicted output

array([1], dtype=int64)

data.iloc[r]["Feedback"]  # actual Output

1.0

from sklearn.metrics import roc_curve, auc

# Compute predicted probabilities
y_pred_prob = rfc.predict_proba(X_test)[:, 1]

# Compute false positive rate, true positive rate, and thresholds
fpr, tpr, thresholds = roc_curve(y_test, y_pred_prob)

# Compute the area under the ROC curve
roc_auc = auc(fpr, tpr)

# Plot the ROC curve
plt.figure(figsize=(10, 8))
plt.plot(fpr, tpr, label="ROC curve (area = %0.2f)" % roc_auc)
plt.plot([0, 1], [0, 1], "k--")  # Random guess line
plt.xlim([0.0, 1.0])
plt.ylim([0.0, 1.05])
plt.xlabel("False Positive Rate")
plt.ylabel("True Positive Rate")
plt.title("Receiver Operating Characteristic (ROC) Curve")
plt.legend(loc="lower right")
plt.show()

import joblib

# Save the model
joblib.dump(rfc, "model.pkl")

['model.pkl']