In [1]:

    

import pandas            as pd
import matplotlib.pyplot as plt
import numpy             as np
import silk
import datetime

%matplotlib inline
pd.options.display.mpl_style='default'
pd.options.display.max_rows=100
pd.options.display.max_columns=30
pd.options.display.width=300


    

In [2]:

    

if (not silk.site.have_site_config()):
    silk.site.init_site('samples/silk.conf')

print "Sensors: ", ', '.join(list(silk.site.sensors()))
print "Classes: ", ', '.join(list(silk.site.classes()))
for cls in silk.site.classes():
    print "Types in class '" + cls + "': ", ', '.join(list(silk.site.types(cls)))
print "Timezone: " + silk.get_configuration("TIMEZONE_SUPPORT")
print "Data Root: " + silk.site.get_data_rootdir()


    

    




Sensors:  S2, S1, S0
Classes:  all
Types in class 'all':  in, out, inweb, outweb, innull, outnull, int2int, ext2ext, inicmp, outicmp, other
Timezone: local
Data Root: /data

In [3]:

    

silkfiles=['samples/1mil.rwz']
print 'loaded {0} files'.format(len(silkfiles))


    

    




loaded 1 files

In [4]:

    

# convenience method to always return a SilkFile
def open_silk(silkfile):
  if silkfile.__class__ == silk.SilkFile:
    return silkfile
  else:
    return silk.silkfile_open(silkfile,silk.READ)

# simple projection of RWRec objects
def map_rwrec(rec):
    return {
        'stime':       rec.stime,
        'application': rec.application,
        'protocol':    rec.protocol,
        'bytes':       rec.bytes,
        'packets':     rec.packets
    }

# lazy iterable projection
def iter_project(silkfile,func=map_rwrec):
  sf=open_silk(silkfile)
  for rec in sf:
    yield func(rec)

# convenience method for creating dataframes from silkfiles
def create_df(silkfile, func=map_rwrec):
    return pd.DataFrame.from_records(iter_project(silkfile),index='stime')


    

In [5]:

    

frame=create_df(silkfiles[0])


    

In [8]:

    

downsample_1hour=frame[['bytes','packets']].resample('1H',how='sum')
downsample_1hour.plot(secondary_y='packets',figsize=(8,8))


    

    
Out[8]:





<matplotlib.axes._subplots.AxesSubplot at 0x107dad450>






    

In [9]:

    

downsample_1minute=frame[['bytes','packets']].resample('1T',how='sum')
downsample_1minute.plot(secondary_y='packets',figsize=(8,8))


    

    
Out[9]:





<matplotlib.axes._subplots.AxesSubplot at 0x107dd8350>






    

In [ ]: