Anshuman Singh

19BCE1352

# import all libraries
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler

data = pd.read_excel("./Multiclass-dataset.xlsx")

data.head()

	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	...	P12	P13	P14	P15	P16	P17	P18	P19	P20	Target Label
0	3.30	7.44	1.52	3.27	0.07	2.14	0.75	0.66	0.0	54.8	...	50.7	6.55	4.09	4.26	0.01	0.00	24.7	2.7	1.6	V1
1	3.43	7.63	1.63	3.27	0.05	2.01	0.74	0.65	0.0	51.8	...	47.9	8.35	5.08	5.01	0.01	0.00	23.3	2.3	1.8	V1
2	3.41	7.32	1.52	3.18	0.07	2.09	0.80	0.70	0.0	54.0	...	54.4	9.27	6.85	7.14	0.19	0.06	25.0	2.5	-0.9	V1
3	3.78	7.85	1.69	3.35	0.03	1.98	0.77	0.67	0.0	57.7	...	48.9	10.26	5.96	5.47	0.05	0.01	24.1	5.6	2.1	V1
4	3.90	7.99	1.61	3.43	0.02	2.14	0.77	0.71	0.0	59.1	...	54.1	8.19	5.81	4.72	0.64	0.16	26.8	2.5	2.1	V1

5 rows × 21 columns

X = data.iloc[:,0:20]

X

	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16	P17	P18	P19	P20
0	3.30	7.44	1.52	3.27	0.07	2.14	0.75	0.66	0.0	54.8	49.7	50.7	6.55	4.09	4.26	0.01	0.00	24.7	2.7	1.6
1	3.43	7.63	1.63	3.27	0.05	2.01	0.74	0.65	0.0	51.8	47.3	47.9	8.35	5.08	5.01	0.01	0.00	23.3	2.3	1.8
2	3.41	7.32	1.52	3.18	0.07	2.09	0.80	0.70	0.0	54.0	50.5	54.4	9.27	6.85	7.14	0.19	0.06	25.0	2.5	-0.9
3	3.78	7.85	1.69	3.35	0.03	1.98	0.77	0.67	0.0	57.7	47.2	48.9	10.26	5.96	5.47	0.05	0.01	24.1	5.6	2.1
4	3.90	7.99	1.61	3.43	0.02	2.14	0.77	0.71	0.0	59.1	54.1	54.1	8.19	5.81	4.72	0.64	0.16	26.8	2.5	2.1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
395	3.78	7.63	1.89	2.91	0.07	1.54	0.82	0.69	0.0	178.1	173.0	147.0	7.78	10.44	11.53	0.02	0.01	73.2	2.9	10.7
396	3.94	7.75	1.78	3.22	0.05	1.81	0.83	0.69	0.0	182.2	183.5	165.6	12.58	13.42	12.30	0.01	0.00	76.3	2.2	5.2
397	3.43	7.39	1.63	3.11	0.05	1.91	0.79	0.68	0.0	171.8	168.4	148.6	12.81	14.65	14.40	0.01	0.00	71.6	2.9	7.6
398	3.96	7.80	1.88	3.03	0.06	1.61	0.82	0.70	0.0	175.0	168.9	144.1	12.03	13.00	11.39	0.00	0.00	71.9	3.2	10.4
399	4.17	8.02	1.86	3.27	0.06	1.75	0.82	0.69	0.0	171.7	169.5	152.8	14.24	16.01	14.09	0.09	0.02	72.0	2.9	5.8

400 rows × 20 columns

y = data[['Target Label']]

y

	Target Label
0	V1
1	V1
2	V1
3	V1
4	V1
...	...
395	V10
396	V10
397	V10
398	V10
399	V10

400 rows × 1 columns

y['Target Label'] = pd.factorize(y['Target Label'])[0]

<ipython-input-98-cb5e4653149a>:1: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  y['Target Label'] = pd.factorize(y['Target Label'])[0]

data.isnull().sum()

P1              0
P2              0
P3              0
P4              0
P5              0
P6              0
P7              0
P8              0
P9              0
P10             0
P11             0
P12             0
P13             0
P14             0
P15             0
P16             0
P17             0
P18             0
P19             0
P20             0
Target Label    0
dtype: int64

data.shape

(400, 21)

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=1)

X_train

	P1	P2	P3	P4	P5	P6	P7	P8	P9	P10	P11	P12	P13	P14	P15	P16	P17	P18	P19	P20
93	3.47	7.56	1.73	3.04	0.05	1.75	0.76	0.66	0.0	61.2	54.7	55.9	10.02	6.63	5.94	0.00	0.00	27.3	3.5	1.5
23	3.26	7.37	1.52	3.18	0.08	2.09	0.75	0.67	0.0	59.4	54.1	54.4	7.46	6.47	6.08	0.50	0.15	26.9	2.7	2.0
299	3.55	7.54	1.61	3.18	0.05	1.98	0.79	0.69	0.0	91.4	59.5	53.1	15.25	8.95	7.46	1.79	0.51	32.9	14.5	10.9
13	3.35	7.42	1.52	3.18	0.05	2.09	0.76	0.69	0.0	49.7	47.5	48.9	6.55	5.40	5.23	0.09	0.03	23.1	1.3	0.8
90	3.65	7.49	1.61	3.18	0.07	1.98	0.82	0.71	0.0	54.7	51.9	54.4	7.74	6.20	6.05	0.00	0.00	25.5	1.9	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
255	3.03	6.91	1.44	3.01	0.04	2.09	0.80	0.70	0.0	114.6	68.5	50.3	11.31	7.24	6.11	2.67	0.88	38.7	19.4	20.2
72	3.90	7.95	1.61	3.43	0.06	2.14	0.78	0.71	0.0	171.4	157.6	128.1	11.83	13.43	13.12	0.04	0.01	68.6	5.1	14.6
396	3.94	7.75	1.78	3.22	0.05	1.81	0.83	0.69	0.0	182.2	183.5	165.6	12.58	13.42	12.30	0.01	0.00	76.3	2.2	5.2
235	3.81	7.97	1.72	3.28	0.06	1.91	0.76	0.68	0.0	88.2	55.8	46.9	20.03	12.10	9.66	0.00	0.00	31.2	14.7	13.1
37	3.60	7.63	1.82	2.96	0.08	1.63	0.78	0.67	0.0	52.3	50.2	50.2	9.66	7.32	6.94	0.32	0.09	24.4	1.0	1.8

320 rows × 20 columns

X_train.columns

Index(['P1', 'P2', 'P3', 'P4', 'P5', 'P6', 'P7', 'P8', 'P9', 'P10', 'P11',
       'P12', 'P13', 'P14', 'P15', 'P16', 'P17', 'P18', 'P19', 'P20'],
      dtype='object')

print(X_train.describe())

               P1          P2          P3          P4          P5          P6  \
count  320.000000  320.000000  320.000000  320.000000  320.000000  320.000000   
mean     3.444937    7.408656    1.594406    3.127719    0.056750    1.973719   
std      0.391488    0.415670    0.136553    0.195338    0.014667    0.191690   
min      2.500000    6.330000    1.250000    2.600000    0.030000    1.220000   
25%      3.167500    7.100000    1.520000    3.010000    0.040000    1.870000   
50%      3.440000    7.420000    1.610000    3.135000    0.060000    1.980000   
75%      3.700000    7.685000    1.660000    3.270000    0.070000    2.090000   
max      4.720000    8.630000    2.210000    3.690000    0.120000    2.580000   

               P7          P8          P9         P10         P11         P12  \
count  320.000000  320.000000  320.000000  320.000000  320.000000  320.000000   
mean     0.787250    0.690500    0.023031  103.526250   81.046250   70.108125   
std      0.026978    0.018205    0.098017   40.757266   44.489461   36.350503   
min      0.670000    0.640000    0.000000   43.300000   42.600000   42.100000   
25%      0.770000    0.680000    0.000000   75.900000   53.600000   49.800000   
50%      0.790000    0.690000    0.000000   93.300000   59.550000   53.100000   
75%      0.810000    0.700000    0.000000  119.475000   77.025000   60.800000   
max      0.870000    0.760000    0.600000  183.200000  183.500000  165.600000   

              P13         P14         P15         P16         P17         P18  \
count  320.000000  320.000000  320.000000  320.000000  320.000000  320.000000   
mean    13.663906   10.147656    8.408312    0.532406    0.156781   39.863750   
std      3.950322    3.302641    3.308026    1.025690    0.303630   16.734755   
min      5.700000    4.090000    3.800000    0.000000    0.000000   20.200000   
25%     10.815000    7.560000    6.057500    0.000000    0.000000   28.800000   
50%     13.650000    9.920000    7.335000    0.010000    0.000000   33.100000   
75%     16.075000   12.415000    9.825000    0.270000    0.082500   41.725000   
max     24.740000   22.180000   20.530000    3.730000    0.990000   76.300000   

              P19         P20  
count  320.000000  320.000000  
mean    10.072813   10.469688  
std      5.917159    6.097812  
min      0.100000   -0.900000  
25%      3.800000    6.400000  
50%     11.350000   10.800000  
75%     15.325000   14.050000  
max     21.800000   26.400000  

y_train.describe()

	Target Label
count	320.000000
mean	4.403125
std	2.867501
min	0.000000
25%	2.000000
50%	4.000000
75%	7.000000
max	9.000000

y_train.head()

	Target Label
93	2
23	0
299	7
13	0
90	2

scaling=StandardScaler()

scaling.fit(X_train)
Scaled_data=scaling.transform(X_train)

principal=PCA(n_components=3)
principal.fit(Scaled_data)
x=principal.transform(Scaled_data)

print(x.shape)

(320, 3)

y['Target Label']

0      0
1      0
2      0
3      0
4      0
      ..
395    9
396    9
397    9
398    9
399    9
Name: Target Label, Length: 400, dtype: int64

principal.components_

array([[ 2.53461730e-01,  2.11796403e-01,  2.24410473e-01,
         1.17530835e-01, -2.05623020e-02, -1.12610573e-01,
         1.42759661e-01,  9.32870155e-02, -6.60062387e-02,
         3.57311566e-01,  3.72636250e-01,  3.67433012e-01,
        -1.17006310e-02,  2.92132738e-01,  3.47494061e-01,
        -3.43828681e-02, -4.21908601e-02,  3.71606307e-01,
        -1.52468751e-01,  9.00843286e-02],
       [-1.05808254e-01, -1.68413902e-01, -1.24655363e-04,
        -2.18960887e-01,  3.57130523e-02, -1.38352480e-01,
         1.99793362e-01,  6.77187983e-02, -9.46563937e-02,
         1.32500716e-01, -1.55627219e-02, -8.39669464e-02,
         3.36260031e-01,  2.25069217e-01, -1.67727086e-02,
         3.76919365e-01,  3.87645351e-01,  4.46358734e-02,
         4.02011856e-01,  4.40246759e-01],
       [-1.71956279e-01, -2.77933150e-01,  2.05352630e-01,
        -4.85272923e-01,  3.07789429e-01, -4.77927155e-01,
         3.39640637e-01, -2.30707871e-01,  2.95676688e-02,
        -3.12664378e-02,  1.75871413e-02,  4.42185780e-02,
        -4.90747874e-02, -2.45163183e-02,  4.28856296e-02,
        -2.00927659e-01, -1.85685973e-01, -6.62953910e-04,
        -1.31392260e-01, -1.50710145e-01]])

plt.figure(figsize=(10,10))
plt.scatter(x[:,0],x[:,1],c=y_train['Target Label'],cmap='plasma')
plt.xlabel('pc1')
plt.ylabel('pc2')

Text(0, 0.5, 'pc2')

# immport relevant libraries for 3d graph
from mpl_toolkits.mplot3d import Axes3D
fig = plt.figure(figsize=(10,10))
  
# choose projection 3d for creating a 3d graph
axis = fig.add_subplot(111, projection='3d')
  
# x[:,0]is pc1,x[:,1] is pc2 while x[:,2] is pc3
axis.scatter(x[:,0],x[:,1],x[:,2], c=y_train['Target Label'],cmap='plasma')
axis.set_xlabel("PC1", fontsize=10)
axis.set_ylabel("PC2", fontsize=10)
axis.set_zlabel("PC3", fontsize=10)

Text(0.5, 0, 'PC3')

print(principal.explained_variance_ratio_)

[0.31884557 0.17022365 0.12875232]

Using Random Forest to make predictions from the result obtatined from PCA

from sklearn.ensemble import RandomForestClassifier

classifier = RandomForestClassifier(max_depth=2, random_state=0)
classifier.fit(X_train, y_train)

y_pred = classifier.predict(X_test)

<ipython-input-138-4b9b744ebb97>:4: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().
  classifier.fit(X_train, y_train)

from sklearn.metrics import confusion_matrix
from sklearn.metrics import accuracy_score

cm = confusion_matrix(y_test, y_pred)
print(cm)
print('\n')
print(f'Accuracy is {accuracy_score(y_test, y_pred)}')

[[6 0 1 0 0 0 0 0 0 0]
 [0 6 0 0 0 0 0 0 0 0]
 [0 0 8 1 0 0 0 0 0 0]
 [0 0 1 2 0 0 0 0 1 0]
 [0 0 0 7 0 0 0 0 0 0]
 [0 0 2 8 2 0 0 0 0 0]
 [0 0 0 0 0 0 7 1 0 0]
 [0 0 0 0 0 0 1 7 0 0]
 [0 0 0 2 0 0 0 0 6 0]
 [0 6 0 0 0 0 0 0 0 5]]


Accuracy is 0.5875

print(f'Accuracy is {accuracy_score(y_test, y_pred)}')

Accuracy is 0.5875

Accuracy is 0.5875

Hence the final outcome of random forest with PCA is 0.5875 or 58.75%