[EMAIL PROTECTED] wrote: > how ? > i have tried to use unix timestamps, and i have also tried with > DateTime objects > do i need to use a scale that isn't linear (default in most) ? > how do i putt this off ?
Here is some code that works for me. It plots multiple datasets against time. The input data looks like this: 2005-04-04 16:00:00 141.154.195.129 - W3SVC1 SP6018ASP2 208.254.37.191 443 GET /rkadqsr/newskills/newskills.cfm selectedTab=1 200 0 53440 599 1594 HTTP/1.1 adqsr.skillport.com Mozilla/4.0+(compatible;+MSIE+6.0;+Windows+NT+5.1) CookiesEnabled=1;+ASPSESSIONIDACTSSRAT=MCNLNLGADPOFGFKAHJHLDDKG;+CFID=785030;+CFTOKEN=27203160 https://adqsr.skillport.com/rkadqsr/login/login.cfm 2005-04-04 16:00:00 208.254.38.216 - W3SVC1 SP6018ASP2 208.254.37.191 443 POST /_rkadqsrBackend_od_cgi/aiccisapi.dll - 200 0 372 274 4547 HTTP/1.0 adqsr.skillport.com aiccisapi - - 2005-04-04 16:00:00 208.254.38.240 - W3SVC1 SP6018ASP2 208.254.37.191 443 POST /_rkadqsrBackend_od_cgi/aiccisapi.dll - 200 0 372 3019 250 HTTP/1.0 adqsr.skillport.com aiccisapi - - import datetime, time import Gnuplot dataPath = 'ex05040416.log' datasets = {} startTime = None f = open(dataPath) for line in f: try: dat, tim, c_ip, cs_username, s_sitename, s_computername, s_ip, s_port, \ cs_method, cs_uri_stem, cs_uri_query, sc_status, sc_win32_status, sc_bytes, \ cs_bytes, time_taken, cs_version, cs_host, cs_User_Agent, cs_Cookie, cs_Referer = line.split() except ValueError: print "Can't parse", line continue tim = time.mktime(time.strptime(dat+' '+tim, "%Y-%m-%d %H:%M:%S")) delay = int(time_taken) if startTime is None: startTime = tim tim -= startTime # print tim, delay datasets.setdefault(sc_status, []).append([tim, delay]) g = Gnuplot.Gnuplot(debug=1) plotter = g.plot for key, values in datasets.items(): plotter(values) plotter = g.replot raw_input('Please press return to continue...\n') -- http://mail.python.org/mailman/listinfo/python-list