How to Customise Pandas Date Time Stamp @ x-axis

Question

When I plots the complete data works fine and displays the date on the x-axis:

.

When I zoom into particular portion to view:

the plot shows only the time rather than date, I do understand with less points can't display different set of date but how to show date or set date format even if the graph is zoomed?

dataToPlot = pd.read_csv(fileName, names=['time','1','2','3','4','plotValue','6','7','8','9','10','11','12','13','14','15','16'], 
                             sep=',', index_col=0, parse_dates=True, dayfirst=True)
dataToPlot.drop(dataToPlot.index[0])
startTime = dataToPlot.head(1).index[0]
endTime = dataToPlot.tail(1).index[0]
ax = pd.rolling_mean(dataToPlot_plot[startTime:endTime][['plotValue']],mar).plot(linestyle='-', linewidth=3,  markersize=9, color='#FECB00')

Thanks in advance!

Answer 1

I have a solution to make the labels look consistent, though bear in mind that it will also include the time on the "larger scale" time plot.

The code below uses the matplotlib.dates functionality to choose a date format for the x-axis. Note that as we're using the matplotlib formatting you can't simple use df.plot but must instead use plt.plot_date and convert your index to the correct format.

import pandas as pd

import matplotlib.pyplot as plt
from matplotlib import dates

# Generate some random data and plot it

time = pd.date_range('07/11/2014', periods=1000, freq='5min')
ts = pd.Series(pd.np.random.randn(len(time)), index=time)

fig, ax = plt.subplots()

ax.plot_date(ts.index.to_pydatetime(), ts.data)

# Create your formatter object and change the xaxis formatting.

date_fmt = '%d/%m/%y %H:%M:%S'

formatter = dates.DateFormatter(date_fmt)
ax.xaxis.set_major_formatter(formatter)

plt.gcf().autofmt_xdate()

plt.show()

An example showing the fully zoomed out plot

An example showing the plot zoomed in.