pca_completed.py

from sklearn.decomposition import PCA
import pandas as pd
from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt


df = pd.read_csv("Iris.csv")
labels = df['Species']
X = df.drop(['Id','Species'],axis=1)

X_std = StandardScaler().fit_transform(X)
  



pca = PCA(n_components=4)
X_transform = pca.fit_transform(X_std)

explained_var = pca.explained_variance_ratio_
for var in explained_var:
	print var
	plt.bar([1,2,3,4],explained_var,label=var)
	plt.xlabel("Component #")
	plt.ylabel("% Variance Contribution")
	plt.legend()
plt.show()

# pca1 = zip(*X_transform)[0]
# pca2 = zip(*X_transform)[1]

# color_dict = {}

# color_dict["Iris-setosa"] = "green"
# color_dict["Iris-versicolor"]='red'
# color_dict["Iris-virginica"] = 'blue'

# i=0
# for label in labels.values:

# 	plt.scatter(pca1[i],pca2[i],color=color_dict[label])
# 	i=i+1

# plt.show()