format xaxis values with time and show them only under the scatter plot value using matplotlib in Python

Question

I am plotting a scatter plot and need help with formatting the xaxis values. Below is my code.

import matplotlib.pyplot as plt
import matplotlib
import numpy as np 
import time
d ={'5000cca234c1c445': {382877: 7, 382919: 3},
'5000cca234c94a2e': {382873: 1, 382886: 1},
'5000cca234c89421': {383173: 1, 383183: 2, 382917: 1, 382911: 1},
'5000cca234c5d43a': {382889: 1, 382915: 1, 382917: 8},
'5000cca234c56488': {382909: 2, 382911: 5}}

colors = list("rgbcmyk")

for data_dict in d.values():
    x = data_dict.keys()
    y = data_dict.values()
    plt.scatter(x,y,color=colors.pop(),s = 60)
    plt.ylabel("Errors" , fontsize=18, color="Green")
    plt.xlabel("Occured at",fontsize=18, color="Green")
    plt.title("DDN44a" , fontsize=25, color="Blue")
    plt.gca().get_xaxis().get_major_formatter().set_useOffset(False)
    plt.xticks(rotation='vertical') 
#plt.ylim(min(y),max(y))
plt.grid()

plt.legend(d.keys())
plt.show()

the x-axis values( 382873,382866,382873.......) are converted dates using "mktime" function by following line of code.

bin = int(mktime(dt) / 3600)

(I am trying to see how many time does the error occur in one hour) On the graph, x-axis values are shown as 382873,382866,382873..etc. I want them to convert them back to the dates(with MM/DD/YY MM:HH:SS FORMAT) they represent. Also I want the date to show directly below the scatter plot (dot) with same color as the dot.

Any help will be greatly appreciated.

Answer 1

You should convert your dates into datetime.datetime objects and use them for your xaxis instead of your number for the time. Use a DateFormatter to get the desired format.

This gives you April 3rd, 2012, 17:30 15 seconds with format:

import datetime
d = datetime.datetime(2012, 4, 3, 17, 30, 15)
print(d.strftime('%m/%d/%y %M:%H:%S'))

04/03/12 30:17:15

DateFormatter takes the same formatting string.

Have a look at annotations for putting the dates below the dots.