How to get the samples in each cluster?

python scikit-learn cluster-analysis k-means
user77005 picture user77005 · Mar 24, 2016 · Viewed 43k times · Source

I am using the sklearn.cluster KMeans package. Once I finish the clustering if I need to know which values were grouped together how can I do it?

Say I had 100 data points and KMeans gave me 5 cluster. Now I want to know which data points are in cluster 5. How can I do that.

Is there a function to give the cluster id and it will list out all the data points in that cluster?

Answer

Praveen picture Praveen · Apr 29, 2017

I had a similar requirement and i am using pandas to create a new dataframe with the index of the dataset and the labels as columns.

data = pd.read_csv('filename')

km = KMeans(n_clusters=5).fit(data)

cluster_map = pd.DataFrame()
cluster_map['data_index'] = data.index.values
cluster_map['cluster'] = km.labels_

Once the DataFrame is available is quite easy to filter, For example, to filter all data points in cluster 3

cluster_map[cluster_map.cluster == 3]

Related questions

Is it possible to specify your own distance function using scikit-learn K-Means Clustering?

Is it possible to specify your own distance function using scikit-learn K-Means Clustering?

Read More
Scikit Learn - K-Means - Elbow - criterion

Today i'm trying to learn something about K-means. I Have understand the algorithm and i know how it works. Now i'm looking for the right k... I found the elbow criterion as a method to detect the right k but …

Read More
Will pandas dataframe object work with sklearn kmeans clustering?

dataset is pandas dataframe. This is sklearn.cluster.KMeans km = KMeans(n_clusters = n_Clusters) km.fit(dataset) prediction = km.predict(dataset) This is how I decide which entity belongs to which cluster: for i in range(len(prediction)): cluster_…

Read More