CODEWITHSUNDEEP
pythontimematplotlibplot
Stefan Bollmann
Stefan Bollmann

Reputation: 720

Is there a ready solution in matplotlib to plot times?

This question has two parts. If it lacks of search for other sources plz be patient. This is part of my problem.

I wrote a script using data produced by tespeed. The data has the format "YYYYMMDDhhmm,down rate, up rate,unit,server" (hh:mm of ...). 

201309221537,0.28,0.04,"Mbit","['speedtest server']"
201309221542,5.78,-1.00,"Mbit","['speedtest server']"
201309221543,0.15,0.06,"Mbit","[...]"

This script plots the above data:

#!/usr/bin/env    
python2.7                                                                                    
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import csv

def main():
    x = []
    y = []
    with open('/path/to/my/public_html/stdout_tespeed_log.csv','r') as csvfile:
        strData = csv.reader(csvfile, delimiter=',')
        for row in strData:
            x += [float(row[0])]
            y += [float(row[1])]
    fig = plt.figure()
    plt.plot(x,y,'+', label='Average download')
    plt.gca().xaxis.major.formatter.set_scientific(False)
    plt.gca().xaxis.major.formatter.set_powerlimits((-2,13))
    locs,labels = plt.xticks()
    plt.xticks(locs, map(lambda x: "%12.0f" % x, locs))
    plt.axis([x[0], x[-1],0,6.5])
    plt.xticks(rotation=20)
    plt.xlabel('Date [YYYYMMDDhhmm]')
    fig.subplots_adjust(bottom=0.2)
    # plt.legend(loc=3)

    plt.gcf().autofmt_xdate()
    plt.savefig("/path/to/my/public_html/speed.png")

main()

At the end this produces a plot like this: the plot

The time axis is not well configured. :-/ The periodically appearing gaps are because of the fact that there are no minutes 60 - 99 in every hour.

Is there some elegant way to accomplish this? Maybe a ready to go module? ;-)

Upvotes: 0

Views: 892

Answers (1)

Veedrac
Veedrac

Reputation: 60167

Matplotlib accepts datetimes, so you can parse the times with

import datetime
datetime.datetime.strptime(row[0], "%Y%m%d%H%M")

and that should work fine.

The formatting options won't work (.set_scientific(False)) this way, though, and your

plt.xticks(locs, map(lambda x: "%12.0f" % x, locs))

should be replaced with something like

import matplotlib.dates as mdates
...
plt.gca().xaxis.major.formatter = mdates.DateFormatter('%Y/%m/%d %H:%M')

Upvotes: 3

Related Questions