histo.py

该软件根据网络数据生成NetFlow记录。NetFlow可用于网络规划、负载均衡、安全监控等

            write('#' + comment)
            write('\n')
        write('# Lower bound ' + fs % (self.lower))
        write('# Upper bound ' + fs % (self.upper))
        write('# Bucket width ' + fs % (self.bucketsz))
        write('# No. buckets ' + '%d' % (self.nbuckets))
        write('# No. samples ' + '%d' % (self.N))
        write('# Total samples ' + '%d' % (self.nsamples))
        if self.N:
            write('# Min. value ' + fs % (self.min))
            write('# Max. value ' + fs % (self.max))
        write('# Min. sample value ' + fs % (self.samplemin))
        write('# Max. sample value ' + fs % (self.samplemax))
        if self.N:
            write('# Mean %f' % mean)
            write('# S.D. %f' % sd)
        write('# Sample mean %f' % samplemean)
        write('# Sample S.D. %f' % samplesd)

        if nlow:
            write('# %d samples below lower bound' % (nlow))
        if nhigh:
            write('# %d samples above upper bound' % (nhigh))
        if self.nlre:
            write('# %d low range exceptions raised' % (self.nlre))
        if self.nhre:
            write('# %d high range exceptions raised' % (self.nhre))
        write('')
        write('# First column is highest inclusive bucket value, second is bucket score')
        write('')

        h = self.ha
        l = self.lower
        bw = self.bucketsz
        for i in range(self.nbuckets):
            count = h[i]
            top = l + ((i+1)*bw)
            if count or zeros:
                write(fsw % (top) + '%10d' % (count))
            hist_vals.append([top, count])
        write('')

        mlist = [(count, top) for top, count in hist_vals]
        if mlist:
            mlist.sort()
            mlist.reverse()
            modes = [mlist[0]]
            for count, top in mlist[1:]:
                if count < modes[0][0]:
                    break
                else:
                    modes.append((count, top))

            if len(modes) == 1:
                write('# Mode %s (%s)' % (bktrange(modes[0][1]), modes[0][0]))
            else:
                modes.reverse()
                ms = '%s (%s)' % (bktrange(modes[0][1]), modes[0][0])
                for count, top in modes[1:]:
                    ms += ', %s (%s)' % (bktrange(top), count)
                write('# Modes %s' % (ms))
        else:
            modes = []

        if draw:
            try:
                from np_data import DataSet, DATA_HIST
                from np_plot import np_Plot, STYLE_BARS, MODE_HIST, MODE_RANGES
            except ImportError, s:
               print 'ImportError \"%s\" - add a directory containing np_data.py and np_plot.py to your PYTHONPATH (try $NPROBE_REPOSITORY_ROOT/analysis/plotter) or cd to $NPROBE_REPOSITORY_ROOT and \'make paths_raw\'' % (str(s))
               sys.exit(1)

            bstring = ''
            sepstr = ''
            if nlow:
                bstring += '%d samples below range' % nlow
                sepstr = ' '
            if nhigh:
                bstring += '%s%d samples above range' % (sepstr, nhigh)
            if bstring:
                bstring = '\n(' + bstring + ' - not included)'

            draw_title = title + bstring + '\n\n' + comment

            ds = DataSet(hist_vals, DATA_HIST, draw_title, 0, style=STYLE_BARS)
            np_Plot([ds], '', style=STYLE_BARS, mode=MODE_HIST|MODE_RANGES, title=title)

        return {'title':title, 'comment':comment, 'lower':self.lower, 'upper':self.upper, 'bucket size':self.bucketsz, 'nbuckets':self.nbuckets, 'nsamples':self.nsamples, 'min':self.min, 'max':self.max, 'samplemin':self.samplemin, 'samplemax':self.samplemax, 'nlow':nlow, 'lows':self.lows, 'nhigh':nhigh, 'highs':self.highs, 'low_range_exceptions':self.nlre, 'high_range_exceptions':self.nhre, 'values':hist_vals, 'mean':mean, 'sample_mean':samplemean, 'sd':sd, 'sample_sd':samplesd, 'modes': modes}
        
	    
##############################################################################
###############################################################################
	
def main():

    """Module histo.py - provides class Histogram with obvious uses.

    When run as a script generates histograms from the specified data file(s).

    Usage: histo.py [flags] [file list]
      Flags:
        The following define operational modes -
          -l<val> Specify lower bound val
          -u<val> Specify upper bound val
          -w<val> Specify bucket width val
          -n<N>   Specify N buckets (default 100)
          -t<N>   Specify auto-ranging based upon first N samples
        The following determine how out of bounds samples are handled -
          default Just count No. of out of range samples
          -o      Save out of range samples
          -r      Throw HistogramLowRangeError and HistogramHighRangeError exceptions
          -e      Catch range error exceptions and extend the histogram accordingly
        The following determine presentation of results -
          -P      Print histogram data at completion
          -D      Fire up plotter showing histogram
          -F<file>Save histogram data to <file>
          -L      Print a dictionary of all histogram data
        Sundry flags -
          -d      Use hard-coded data set for testing
          -f<n>   Take samples from field n of input data file (default 1)
          -v      Verbose mode
          -h      This help
          -H      Help on the Histogram class itself

      Auto-ranging:
         If lower and upper histogram bounds are specified all samples are
           immediately added in, and out of range samples dealt with as
           determined by the -o, -r, and -e flags.
         If neither, or only one bound, are specified the sample values are
           stored until a trigger number of samples - specified using
           the -t flag - have been presented, at which point an upper/lower
           bound, or both, are automatically calculated based upon the current
           max/min stored sample values, and the stored values are added in.
           Bucket widths or nbuckets are calculated based upon the specified
           nbuckets or width - if neither are specified the default
           nbuckets = 100 is used.
         The default trigger value of 0 results in all samples being stored and
         auto range calculations being performed on the entire sample set at
         completion.
        
      """
    
    def iof(s):

        if s.find('.') >= 0:
            return float(s)
        else:
            return int(s)

    import os
    from sys import argv
    import getopt
    try:
        from np_data import DataReader
    except ImportError, s:
       print 'ImportError \"%s\" - add a directory containing np_data.py to your PYTHONPATH (try $NPROBE_REPOSITORY_ROOT/analysis/plotter) or cd to $NPROBE_REPOSITORY_ROOT and \'make paths_raw\'' % (str(s))
       sys.exit(1)

    data_def =  [0, 1,2,2,3,4,5,6,7,8,9, 10, 9.5, 6.000001, 11]
    #data_def =  [0,0, 1,2,3,4,4,5,6,7,8,9,10]

    scriptname = os.path.basename(argv[0])
    
    try:
	optlist, args = getopt.getopt(sys.argv[1:], 'l:u:n:w:t:oPF:LDdervf:hH')

    except getopt.error, s:
        print '%s: %s' % (scriptname, s)
	sys.exit(1)

    lower = upper = nbuckets = bsz = ntrig = orange = None
    printit = list = draw = throw_e = extend = verbose = 0
    get_data = field = 1
    ofile = ''
    stuff = None

    for opt in optlist:
        if opt[0] == '-l':
            lower = iof(opt[1])
        if opt[0] == '-u':
            upper = iof(opt[1])
        if opt[0] == '-n':
            nbuckets = iof(opt[1])
        if opt[0] == '-w':
            bsz = iof(opt[1])
        if opt[0] == '-t':
            ntrig = iof(opt[1])
        if opt[0] == '-o':
             orange = 1
        if opt[0] == '-P':
            printit = 1
        if opt[0] == '-L':
            list = 1
        if opt[0] == '-F':
            ofile = opt[1]
        if opt[0] == '-D':
            draw = 1
        if opt[0] == '-d':
            get_data = 0
        if opt[0] == '-r':
            throw_e = 1
        if opt[0] == '-e':
            extend = 1
        if opt[0] == '-v':
            verbose = 1
        if opt[0] == '-f':
            field = int(opt[1])
        if opt[0] == '-H':
            h = Histogram()
            print
            print h.__doc__
            sys.exit(0)
        if opt[0] == '-h':
            print
            print main.__doc__
            sys.exit(0)

    histos = []

    if get_data or args:
        dr = DataReader(args)
        stuff = dr.get_data()

    if not stuff:
        title = 'Test set'
        data = [data_def]
    else:
        title=stuff[1]
        data = []
        for d in stuff[0]:
            data.append(d.data)
            
    #print data

    for d in data:
        try:
            h = Histogram(lower=lower, upper = upper, nbuckets= nbuckets, bucketsz=bsz, ntrigvals=ntrig, save_orange=orange, rangeexcept=throw_e, verbose=verbose)
            t = type(d[0])
            if t is IntType or t is FloatType or t is ListType or t is TupleType:
                if t is ListType or t is TupleType:
                    try:
                        d = [t[field] for t in d]
                    except IndexError:
                        print 'Specified field %d not found in data set - quitting' % (field)
                        sys.exit(1)
                for v in d:
                    try:
                        h.add(v)
                    except HistogramHighRangeError, s:
                        if verbose:
                            print s
                        if extend:
                            h.extend(v)
                        else:
                            print 'Value', v, 'above bound %s - quitting' % (str(h.upper))
                            sys.exit(1)
                            
                    except HistogramLowRangeError, s:
                        if verbose:
                            print s
                        if extend:
                            h.extend(v)
                        else:
                            print 'Value', v, 'below or equals bound %s - quitting' % (str(h.lower))
                            sys.exit(1)
            else:
                print scriptname, 'ERROR: data points appear to be', t
                sys.exit(1)

            histos.append(h)
        except HistogramError, s:
            print s
            return

    for h in histos:
        d = h.results(printit=printit, file=ofile, draw=draw, title=title, comment='')
        if list:
            d = d.items()
            d.sort()
            for i in d:
                if type(i[1]) is ListType and i[1]:
                    print '%-10s' % (i[0])
                    for v in i[1]:
                        print '\t', v
                elif type(i[1]) is StringType:
                    print '%-10s \'%s\'' % (i[0], i[1])
                else:
                    print '%-10s' % (i[0]), i[1]
        
	    
##############################################################################
###############################################################################


# Call main when run as script
if __name__ == '__main__':
        main()