Adding labels to a matplotlib graph

Question

Having the following code:

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.dates as mdates

days, impressions = np.loadtxt('results_history.csv', unpack=True, delimiter=',',usecols=(0,1) ,
        converters={ 0: mdates.strpdate2num('%d-%m-%y')})

plt.plot_date(x=days, y=impressions, fmt="r-")
plt.title("Load Testing Results")


#params = {'legend.labelsize': 500,
    #'legend.handletextpad': 1,
    #'legend.handlelength': 2,
    #'legend.loc': 'upper left',
    #'labelspacing':0.25,
    #'legend.linewidth': 50}
#plt.rcParams.update(params)

plt.legend("response times")

plt.ylabel("Date")
plt.grid(True)
plt.show()

The graph is generated but i can't figure how can i add some xy labels. The generated graph:

Also tried to increase the legend text size but the text is not displayed. And the labels from the X axis are overlapped. CSV file:

01-05-14, 55494, Build 1
10-05-14, 55000, Build 2
15-05-14, 55500, Build 3
20-05-14, 57482, Build 4
25-05-14, 58741, Build 5

How can i add the xytext from the csv and also change the format for legend and X axis?

Solution

You need annotate, e.g.:

plt.annotate('some text',xy=(days[0],impressions[0]))

To adjust the x axis text you could add:

fig=plt.figure() # below the import statements
...
fig.autofmt_xdate() # after plotting

To change the legend text use the label parameter in your plot function:

plt.plot_date(x=days, y=impressions, fmt="r-",label="response times")

To increase the legend font size do this:

plt.legend(fontsize='x-large')