10/14/23, 12:20 PM BHMC17 p5.

ipynb - Colaboratory

import numpy as np # linear algebra

import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
import seaborn as sns #visualization
import matplotlib.pyplot as plt
import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename))

data = pd.read_csv("/EconomiesOfScale.csv")

from google.colab import drive

drive.mount('/content/drive')

Mounted at /content/drive

data.head()

output Number of Units Manufacturing Cost

0 1.000000 95.066056

1 1.185994 96.531750

2 1.191499 73.661311

3 1.204771 95.566843

4 1.298773 98.777013

data.describe()

Number of Units Manufacturing Cost

count 1000.000000 1000.000000

mean 4.472799 40.052999

std 1.336241 10.595322

min 1.000000 20.000000

25% 3.594214 32.912036

50% 4.435958 38.345781

75% 5.324780 44.531822

max 10.000000 100.000000

nans=pd.isnull(data).sum()
nans[nans>0]

Series([], dtype: int64)

data.shape[0]

1000

Let's carry out some visualizations using Seaborn

ax = sns.boxplot(x=data["Manufacturing Cost"])

https://colab.research.google.com/drive/1Uvk4Nqweukf7QH3oYbfyTY417obmtB4J#printMode=true 1/4
10/14/23, 12:20 PM BHMC17 p5.ipynb - Colaboratory

ax = sns.boxplot(x=data['Number of Units'])

sns.jointplot(data=data, x="Number of Units", y="Manufacturing Cost", kind="reg")

<seaborn.axisgrid.JointGrid at 0x7cfbc13ba6b0>

from sklearn.preprocessing import MinMaxScaler

sc = MinMaxScaler(feature_range = (0, 1))
scaled_data = sc.fit_transform(data)
scaled_data

array([[0. , 0.9383257 ],
[0.02066596, 0.95664687],
[0.02127763, 0.67076638],
...,
[0.86454312, 0.07467234],

https://colab.research.google.com/drive/1Uvk4Nqweukf7QH3oYbfyTY417obmtB4J#printMode=true 2/4
10/14/23, 12:20 PM BHMC17 p5.ipynb - Colaboratory
[0.87752219, 0.06422889],
[1. , 0.01934721]])

scaled_data = pd.DataFrame(scaled_data, columns = ['Number of Units', 'Manufacturing Cost'])

scaled_data

Number of Units Manufacturing Cost

0 0.000000 0.938326

1 0.020666 0.956647

2 0.021278 0.670766

3 0.022752 0.944586

4 0.033197 0.984713

... ... ...

995 0.788857 0.048188

996 0.859972 0.094207

997 0.864543 0.074672

998 0.877522 0.064229

999 1.000000 0.019347

1000 rows × 2 columns

y = scaled_data.pop('Manufacturing Cost')
y

0 0.938326
1 0.956647
2 0.670766
3 0.944586
4 0.984713
...
995 0.048188
996 0.094207
997 0.074672
998 0.064229
999 0.019347
Name: Manufacturing Cost, Length: 1000, dtype: float64

X = scaled_data.values

"""
Split the dataset into training set and test set with an 80-20 ratio
"""
from sklearn.model_selection import train_test_split
seed=1
X_train, X_test, \
y_train, y_test = train_test_split(X, y, test_size=0.2, \
random_state=42)

from sklearn.svm import SVR

regr= SVR(C=1.0, epsilon=0.2)

regr.fit(X_train,y_train.ravel())

▾ SVR
SVR(epsilon=0.2)

y_pred = regr.predict(X_test)
y_pred

array([0.23476998, 0.19564829, 0.19552012, 0.20350118, 0.25761209,

0.20152073, 0.20868388, 0.23718671, 0.20304453, 0.35180452,
0.19570875, 0.41292374, 0.20709024, 0.23183355, 0.22813452,
0.21300845, 0.28898906, 0.20898637, 0.19462914, 0.27781544,
0.22914077, 0.33378322, 0.26540822, 0.23374456, 0.317575 ,
0.30691577, 0.38566758, 0.22948237, 0.21358697, 0.22992041,
0.27087766, 0.35108264, 0.20962335, 0.24059563, 0.26569672,
0.32181694, 0.20065724, 0.22099506, 0.21344844, 0.45573471,
0.27529935, 0.38878804, 0.27968523, 0.23177926, 0.28450419,
0.29023651, 0.22285173, 0.2161231 , 0.25313191, 0.19790806,
0.2250116 , 0.20793397, 0.29438255, 0.22256105, 0.60464413,

https://colab.research.google.com/drive/1Uvk4Nqweukf7QH3oYbfyTY417obmtB4J#printMode=true 3/4
10/14/23, 12:20 PM BHMC17 p5.ipynb - Colaboratory
0.58535064, 0.21132973, 0.70114225, 0.31394558, 0.19646868,
0.28550935, 0.46230228, 0.23054964, 0.3178442 , 0.21526703,
0.20025726, 0.20334417, 0.2047899 , 0.41879638, 0.23031656,
0.37798776, 0.24097249, 0.22005137, 0.19537004, 0.28626906,
0.28661788, 0.20441796, 0.5164567 , 0.22013279, 0.28129508,
0.20147005, 0.42688768, 0.29052581, 0.42638852, 0.27804957,
0.22929735, 0.22977071, 0.19732925, 0.19497802, 0.27538463,
0.2633768 , 0.25082279, 0.23453376, 0.31433468, 0.19467478,
0.25353211, 0.19456455, 0.39709234, 0.43353714, 0.19685745,
0.1971824 , 0.26981381, 0.22775389, 0.29543255, 0.20733488,
0.20571743, 0.22817938, 0.23200193, 0.33702665, 0.22039579,
0.23372122, 0.19661119, 0.22950817, 0.19593826, 0.39894177,
0.25449299, 0.32918061, 0.23097683, 0.2329515 , 0.28591771,
0.2110974 , 0.25549309, 0.19736443, 0.29471497, 0.30469764,
0.22748822, 0.22111173, 0.25231059, 0.2048678 , 0.19524356,
0.21631618, 0.20047142, 0.29288796, 0.22982797, 0.23354053,
0.46075801, 0.36576062, 0.31622165, 0.59955219, 0.41707837,
0.49931476, 0.29957463, 0.19680618, 0.28884013, 0.21141338,
0.21829503, 0.19566166, 0.30106666, 0.19659288, 0.27753212,
0.20592251, 0.34307049, 0.23110011, 0.25478658, 0.22832641,
0.24935498, 0.28069803, 0.19986757, 0.45245932, 0.20906462,
0.21711763, 0.19448755, 0.20520075, 0.205105 , 0.20446054,
0.20282835, 0.2849949 , 0.20061093, 0.27949631, 0.28022113,
0.26771314, 0.24630999, 0.37452241, 0.2290563 , 0.2430412 ,
0.26757145, 0.29705325, 0.2948896 , 0.19535961, 0.20187177,
0.24093193, 0.20266503, 0.26391159, 0.2341701 , 0.19928893,
0.35138871, 0.26921077, 0.26709635, 0.22872646, 0.19508277,
0.23028492, 0.19709374, 0.20392945, 0.32127952, 0.31694141,
0.25830457, 0.27435644, 0.31842808, 0.21207435, 0.40956217])

from sklearn.metrics import r2_score,mean_squared_error

mse = mean_squared_error(y_test,y_pred)
rmse= np.sqrt(mse)
rmse

0.0822903295452384

mse

0.006771698336663936

r2_score(y_test, y_pred)

0.4700318323625766

The R-square value is approximately 0.5 which is quite encouraging. We can improve this work by
trying out algorithms as well.

https://colab.research.google.com/drive/1Uvk4Nqweukf7QH3oYbfyTY417obmtB4J#printMode=true 4/4