Plotting datetime output using matplotlib

Question

So I have this code based on a simple data array that looks like this:

    5020 : 2015 7 11 11 42 54 782705
    5020 : 2015 7 11 11 44 55 575776
    5020 : 2015 7 11 11 46 56 560755
    5020 : 2015 7 11 11 48 57 104872

and the plot looks like the following:

    import scipy as sp
    import matplotlib.pyplot as plt
    data = sp.genfromtxt("E:/Python/data.txt", delimiter=" : ")
    x = data[:,0]
    y = data[:,1]
    plt.scatter(x,y)
    plt.title("Instagram")
    plt.xlabel("Time")
    plt.ylabel("Followers")
    plt.xticks([w*2*60 for w in range(10)],
    ['2-minute interval %i'%w for w in range(10)])
    plt.autoscale(tight=True)
    plt.grid()
    plt.show()

I'm looking for a simple way to use the datetime output as x intervals on the graph, I can't figure out a way to make it understand it and there's this:

    In [15]:sp.sum(sp.isnan(y))
    Out[15]: 77

Which I guess is because of the spaces? I'm new to machine learning in Python, forgive my ignorance.

Thank you very much.

Sasha · Accepted Answer · 2015-07-11 18:50:39Z

1

I would solve this by directly passing datetime.datetime objects to pyplot. Here is a short example:

import datetime as dt
import matplotlib.pyplot as plt
import matplotlib

# Note: please figure out yourself the data input
x     = [dt.datetime(2015,7,11,11,42,54),
     dt.datetime(2015,7,11,11,44,56),
     dt.datetime(2015,7,11,11,46,56),
     dt.datetime(2015,7,11,11,48,57)]

#define the x limit:
xstart= dt.datetime(2015,7,11,11,40,54)
xstop = dt.datetime(2015,7,11,11,50,54)


y     = [782705, 575776, 560755, 104872]

fig,ax= plt.subplots()
ax.scatter(x,y)
xfmt = matplotlib.dates.DateFormatter('%D %H:%M:%S')
ax.xaxis.set_major_formatter(xfmt)
ax.set_title("Instagram")
ax.set_xlabel("Time")
ax.set_ylabel("Followers")
ax.set_xlim(xstart,xstop)
plt.xticks(rotation='vertical')
plt.show()

Result: enter image description here

answered Jul 11, 2015 at 18:50

Sasha

1,4182 gold badges14 silver badges22 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Leb · Accepted Answer · 2015-07-11 18:59:22Z

0

Yes it's because of the spaces. When you're importing the data it's assigning NaN to your x values.

Try this, it's a little longer but should work:

data = []
x=[]
y=[]

with open('data.txt', 'r') as f:
    for line in f:
        data.append(line.split(':'))

for i in data:
y.append(i[0])
x_old.append(i[1])

for t in x_old:
    x.append(float(t[17:19]+'.'+t[20:])/60+int(t[14:16]))

Because of the spaces I had to convert the data into float manually. I divided the seconds+milliseconds by 60 then added to minutes since I'm assuming you're only interested in that (2 min interval).

If the format is done better you can use datetime and extract the information better. For example:

my_time = datetime.strptime('2015 7 11 11 42 54.782705', '&Y &m %d %H:%M:%S.%f')

edited Jul 11, 2015 at 18:59

answered Jul 11, 2015 at 18:18

Leb

16k11 gold badges58 silver badges78 bronze badges

2 Comments

Alloy Over a year ago

I tried this and got: ValueError: invalid literal for float(): 2015 7 11 11 42 54 782705 EDIT: it's because of a non printable char '\n'

Leb Over a year ago

Ok then, I edited my post thinking it was another conversion factor. Glad it worked.

Collectives™ on Stack Overflow

Plotting datetime output using matplotlib

2 Answers 2

Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related