def k_means(points, k, iterations=10):
"""
K Means Unsupervised ML Algorithm Implementation with Forgy Initialization
Input:
points (numpy array): a 2D (M,N) Array with the data to cluster.
M = Instances, N = variables.
k (int): The number of clusters to find
"""
idx = np.random.randint(len(points),size=k)
centroids = points[idx,:]
clusters = get_clusters(points,centroids)
for i in range(iterations):
# Use get_clusters() and update_centroids() for each iteration.
return clusters,centroids
if __name__ == "__main__":
data = load_data('./data/iris.data')
k = 3
X = np.array([f[:-1] for f in data])
y = np.array([f[-1] for f in data])
clusters,centroids = k_means(X,3)
show_clusters_centroids(clusters, centroids, "Result", keep=True)
def k_means(points, k, iterations=10):
"""
K Means Unsupervised ML Algorithm Implementation with Forgy Initialization
Input:
points (numpy array): a 2D Array with the data to cluster.
k (int): The number of clusters to find
"""
idx = np.random.randint(len(points),size=k)
centroids = points[idx,:]
clusters = get_clusters(points,centroids)
for i in range(iterations):
if i % 1 == 0:
if i == 0:
title = "Initialization"
else:
title = "Iteration {}".format(i+1)
show_clusters_centroids(
clusters,
centroids,
title,
)
clusters = get_clusters(points,centroids)
centroids = update_centroids(clusters)
return clusters,centroids
def k_means(puntos):
cantidad_de_centros = 0
points = np.array(puntos)
cantidad_de_centros = int(input("Elija la cantidad de centros: "))
system('clear')
iterationts = int(input("Elija la cantidad de iteraciones: "))
idx = np.random.randint(len(points), size=cantidad_de_centros)
k_lista = points[idx, :]
clusters = cercanos(points, k_lista)
# Redifine puntos and centers adjusting with the methods cercanos() and centros()
for i in range(iterationts):
if i % 1 == 0:
if i == 0:
title = "Initialization"
else:
title = "Iteration {}".format(i + 1)
show_clusters_centroids(clusters, k_lista, title)
clusters = cercanos(points, k_lista)
k_lista = centros(clusters)
# Return the new adjustes list
return clusters, k_lista
def main():
filename = './iris.data'
data = load_data(filename)
#consigue todas las columnas, menos la ultima
X = np.array([f[:-1] for f in data])
#consigue la ultima columna, de nombres
y = np.array([f[-1] for f in data])
#inicializa k_means
clusters, centros = k_means(X)
#imprime los valores de los centros, en 4 dimensiones
print(centros)
#genera una grafica con los valores finales
show_clusters_centroids(clusters, centros, "Result", keep=True)
#muestra la grafica
plt.show()
def k_means(puntos, k, iterations=10):
idx = np.random.randint(len(puntos), size=k)
centros = puntos[idx, :]
clusters = cercanos(puntos, centros)
for i in range(iterations):
show_clusters_centroids(
clusters,
centros,
"Graph",
)
clusters = cercanos(puntos, centros)
centros = find_centros(clusters)
return clusters, centros
def k_means(puntos, k, iterations=10):
idx = np.random.randint(len(puntos), size=k)
centros = puntos[idx, :]
clusters = cercanos(puntos, centros)
for i in range(iterations):
show_clusters_centroids(
clusters,
centros,
"Graph",
)
clusters = cercanos(puntos, centros)
centros = find_centros(clusters)
return clusters, centros
if __name__ == "__main__":
data = load_data('./data/iris.data')
k = 4
X = np.array([f[:-1] for f in data])
y = np.array([f[-1] for f in data])
clusters, centros = k_means(X, k)
show_clusters_centroids(clusters, centros, "Result", keep=True)