Recommendation - Distance Measures

 

Ref - Link

	#cosine similarity example python
	#importing libraries
	from sklearn.metrics.pairwise import cosine_similarity
	from scipy import spatial
	#creating two vectors
	vector_a = [1,1,0,0]
	vector_b = [1,1,1,0]
	vector_c = [1,0,1,0]
	#calculating cosine similarity
	cos_sim = cosine_similarity([vector_a], [vector_b])[0][0]
	#printing cosine similarity and cosine distance
	print('Cosine similarity: A - B ', cos_sim)
	#calculating cosine similarity
	cos_sim = cosine_similarity([vector_a], [vector_c])[0][0]
	#printing cosine similarity and cosine distance
	print('Cosine similarity: A - C ', cos_sim)
	#Euclidean similarity example
	#importing the necessary libraries
	from scipy.spatial import distance
	#calculating the euclidean distance
	euclidean_distance = distance.euclidean(vector_a, vector_b)
	#calculating the euclidean similarity
	euclidean_similarity = 1/(1 + euclidean_distance)
	#printing the euclidean similarity
	print(euclidean_similarity)
	#calculating the euclidean distance
	euclidean_distance = distance.euclidean(vector_a, vector_c)
	#calculating the euclidean similarity
	euclidean_similarity = 1/(1 + euclidean_distance)
	#printing the euclidean similarity
	print(euclidean_similarity)
	#manhattan similarity example
	# Calculate Manhattan similarity between two vectors
	import numpy as np
	def manhattan_similarity(vec1, vec2):
	return np.sum(np.abs(vec1 - vec2))
	print(manhattan_similarity(np.array(vector_a), np.array(vector_b)))
	print(manhattan_similarity(np.array(vector_a), np.array(vector_c)))

view raw similaritydistances.py hosted with ❤ by GitHub

Keep Exploring!!!