Chapter 4: Dimension Reduction and PCA

Activity 6: Manual PCA versus scikit-learn

Solution

1. Import the pandas, numpy, and matplotlib plotting libraries and the scikit-learn PCA model:

   import pandas as pd
   import numpy as np
   import matplotlib.pyplot as plt
   from sklearn.decomposition import PCA

2. Load the dataset and select only the sepal features as per the previous exercises. Display the first five rows of the data:

   df = pd.read_csv('iris-data.csv')
   df = df[['Sepal Length', 'Sepal Width']]
   df.head()

   The output is as follows:

   Figure 4.43: The first five rows of the data

3. Compute the covariance matrix for the data:

   cov = np.cov(df.values.T)
   cov

   The output is as follows:

   Figure 4.44: The covariance matrix for the data

4. Transform the data using the scikit-learn API and only the first principal component. Store the transformed data in the sklearn_pca variable:

   model = PCA(n_components=1)
   sklearn_pca = model.fit_transform(df.values)

5. Transform the...