I live in Germany and Germany is known for its diversity. The people are diverse, the culture is diverse and the quality of internet connections are equally as diverse. Having spoken to my internet provider who insisted I had no problems with my internet connection I decided to write a program to monitor my upload and download speed.
Getting started
We will be using two cool packages in python namely speedtest and matplotlib. speedtest will provide us with the information about our internet connection and matplotlib will allow us to generate graphs based on this information.
Whether you're in a virtual environment or not using one you will have to install speedtest and matplotlib.
pip install matplotlib
pip install speedtest-cli
We will also be using a few built in python packages datetime and csv. CSV (or comma separated values) are a quick way of storing data. We will be using this to store the information about our internet speeds and then using matplotlib to make this information visual.
Gathering data
Create a python file called monitor.py
touch monitor.py
and let's put the following code on the inside
import speedtest
s = speedtest.Speedtest()
while True:
print(s.download(), s.upload())
and we should be greeted by the following output before hitting ctrl + C to get out of this infinite loop.
75020309.22943133 24381170.373616524
105192450.00822952 40653433.153288215
So by creating a new instance of speedtest as s and testing the upload and download speed we are given the upload and download speed in bits per second. To convert this to megabits per second (Mb/s) we can do the following to include the time of the test too:
import speedtest
import datetime
s = speedtest.Speedtest()
while True:
time = datetime.datetime.now().strftime("%H:%M:%S")
downspeed = round((round(s.download()) / 1048576), 2)
upspeed = round((round(s.upload()) / 1048576), 2)
print(f"time: {time}, downspeed: {downspeed} Mb/s, upspeed: {upspeed} Mb/s")
which yields:
time: 12:44:15, downspeed: 95.04 Mb/s, upspeed: 32.85 Mb/s
time: 12:44:35, downspeed: 99.46 Mb/s, upspeed: 38.76 Mb/s
time: 12:44:56, downspeed: 100.59 Mb/s, upspeed: 38.94 Mb/s
Now we will move on to recording this in a CSV file. CSVs are large text files which values separated by commas. Here is an example of one of mine:
time,downspeed,upspeed
12:17:01,100.05,38.28
12:17:21,100.53,37.85
12:17:42,74.87,25.92
We start with a header row then have the information split up. We can think about this like an excel sheet:
In order to record into a csv file in python we need to import the CSV package and 'open' a CSV file (if one doesn't exist it will create one).
import speedtest
import datetime
import csv
s = speedtest.Speedtest()
with open('test.csv', mode='w') as speedcsv:
csv_writer = csv.DictWriter(speedcsv,
fieldnames=['time', 'downspeed', 'upspeed'])
csv_writer.writeheader()
while True:
time = datetime.datetime.now().strftime("%H:%M:%S")
downspeed = round((round(s.download()) / 1048576), 2)
upspeed = round((round(s.upload()) / 1048576), 2)
csv_writer.writerow({
'time': time,
'downspeed': downspeed,
"upspeed": upspeed
})
So while you let this code run for 4-5 minutes we can discuss what is going on. Line 7 with open essentiallly creates a csv file with the name test.csv with the headers name, downspeed and upspeed and writers them into the csv. Then the loop begins and every time a test is performed by speedtest it writes a new row into the csv with the time, download speed and upload speed we specified before. So let's go and look at that now.
time,downspeed,upspeed
12:51:16,99.29,38.66
12:51:37,100.67,38.79
12:51:57,99.7,38.79
12:52:17,92.89,31.99
12:52:38,99.4,38.96
Cool now we are gathering information and storing it in a csv. You could do all sorts of clever stuff with the filename like name it today's date using datetime but for now I will keep it as test.
You could also set up a check to see if a certain amount of time has passed and to stop the application after that but I will leave that up to your own creativity.
Visualising the data
Let's make another python file to generate the graph of our internet connection. This is where we will use matplotlib.
touch make_graph.py
import matplotlib.pyplot as plt
import csv
import matplotlib.ticker as ticker
times = []
download = []
upload = []
The easiest way to generate a graph is by using an array. In order to populate our arrays we will have to iterate through our new csv file
with open('test.csv', 'r') as csvfile:
plots = csv.reader(csvfile, delimiter=',')
next(csvfile)
for row in plots:
times.append(str(row[0]))
download.append(float(row[1]))
upload.append(float(row[2]))
print(times, "\n", download, "\n", upload)
['12:51:16', '12:51:37', '12:51:57', '12:52:17', '12:52:38']
[99.29, 100.67, 99.7, 92.89, 99.4]
[38.66, 38.79, 38.79, 31.99, 38.96]
So now we are parsing our data! The next(csvfile) essentially skipss the row of headers (that were for our benefit only, not python's). Now we come on to using matplotlib which I am by no standards and expert on. Their documentation is extensive.
plt.figure(30, 30)
plt.plot(times, download, label='download', color='r')
plt.plot(times, upload, label='upload', color='b')
plt.xlabel('time')
plt.ylabel('speed in Mb/s')
plt.title("internet speed")
plt.legend()
plt.savefig('test_graph.jpg', bbox_inches='tight')
This code tells matplotlib or plt to create a figure with the lines labelled 'download' and 'upload' to be plotted against their respective arrays. So for every item in the download and and time array a mark will be made on the graph against it. Finally the x and y axis get labels and the graph gets saved.
Potential improvements
This is a short form of a project I recently finished which contains automated uploads to a cloud platform, schedules itself daily does some annotations of the lowest download speed. I initially made it to complain to my internet provider however I know that it will be like showing a graph to a brick wall.
If you have any questions about this tutorial please don't hesitate to reach out to me.
Top comments (7)
I just started playing around with Python, and this seems like a pretty cool weekend project! Thanks for sharing.
Very nice tutorial. ISPs are generally very hard to solve problems and you have found a good solution and also found a good starter project !
Just sharing a info based on my experience reproducing this tutorial with python 3.8.
There is a argument that should be added in monitor.py:
with open ('./test.csv', mode='w', newline='') as speedcsv:
Without this argument each iteration will add a line with data and a blank line after. When make_graph.py read test.csv file the method that reads and append to the arrays (times,download and upload) will cause a "out of index exception".
Best regards and thanks for sharing !
that's smart! I have a weird ISP only for the benefit of having an upload speed around the same as my download, but the internet speeds change a lot during the day
also the plt.figure(30, 30) line should be:
plt.figure('speedtest', [30, 30])
This could be a good starter project thanks for sharing.
I would add a line:
speedcsv.flush()
because it was not writing to the file. Nice script man.
Can we also correlate this data with multiple devices around us, I have read that the main cause of fluctuation is due to inteference of frequencies?