[scikit-learn] Clustering information from one file

[Fabio Santos]

Hi all,
My name is Fábio and I'm new in scikit, and I trying to cluster information from one file with python script (i fount on web). But i saw that the output had problem with numbers...See:
Script#
import clickimport reimport numpyimport random
from collections import defaultdict
from sklearn.feature_extraction.text import TfidfVectorizerfrom sklearn.cluster import KMeans

@click.command()@click.argument('filename')@click.option('--clusters', default=50, help='Number of clusters')@click.option('--sample', default=400, help='Number of samples to print')def cluster_lines(filename, clusters, sample):    lines = numpy.array(list(_get_lines(filename)))
    doc_feat = TfidfVectorizer().fit_transform(lines)    km = KMeans(clusters).fit(doc_feat)
    k = 0    clusters = defaultdict(list)    for i in km.labels_:      clusters[i].append(lines[k])      k += 1
    s_clusters = sorted(clusters.values(), key=lambda l: -len(l))
    for cluster in s_clusters:        print 'Cluster [%s]:' % len(cluster)        if len(cluster) > sample:            cluster = random.sample(cluster, sample)        for line in cluster:            print line        print '--------'

def _clean_line(line):    line = line.strip().lower()    line = re.sub('\d+', '(N)', line)    return line

def _get_lines(filename):    for line in open(filename).readlines():        yield _clean_line(line)

if __name__ == '__main__':    cluster_lines()
output [root at vmcaiosyscolprod01 71001492]#  python Cluster-LearnMachine.py DataSets/ospf.teste3Cluster [7]:"rjbotaa max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area""rjmteab max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area""rjmckaa max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area""rjdqcaa max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area""rjdqcab max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area""rjcenaa max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area""rjcenab max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area"--------Cluster [1]:"rjbotab max-metric router-lsa on-startup log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area"--------Cluster [1]:"rjmteaa ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area"--------Cluster [1]:"rjmckab max-metric router-lsa on-startup ispf log-adjacency-changes detail auto-cost reference-bandwidth timers throttle spf timers throttle lsa timers lsa arrival timers pacing flood passive-interface default maximum-paths mpls ldp sync mpls traffic-eng router-id loopback mpls traffic-eng area"--------

See that the output shown (N) on numbers,  and i'm not fount a way to use the big cluster as a template fo fount diference between the bigger cluster and others clusters. How can i do that?
Thanks

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/scikit-learn/attachments/20161209/cbf55b6d/attachment-0001.html>