how to plot data selectively in python

Question

I'm trying to do some data analyzing using python and I have got the data displayed successfully. But there is a problem with the figure.

As you can see, the x-axis(date) is quite messed up. The reason is that the data format is like(collected on daily-basis):

2014-12-30,1423.98,1424.16,1407.51,1407.51
2014-12-29,1433.80,1433.84,1409.53,1424.67
...

I put the first row(dates) in an array and used the array as x-axis.

I have a solution myself, which is to only output x-axis when the month changes(like '2014-11', '2014-12',...), but I don't know any function that can do this...

Can anyone help me with my problem? Thanks!

Please be specific about the tools you are using and your current attempt. What kind of answer can you expect to your current question? Note that you are the only person who knows what code you have written. — 5gon12eder, Nov 27 '15 at 02:23
usually its possible to set the number of vertical lines to be drawn, called ticks in matplotlib for example: http://stackoverflow.com/questions/6682784/how-to-reduce-number-of-ticks-with-matplotlib — Emile Vrijdags, Nov 27 '15 at 02:24

score 1 · Answer 1 · answered Nov 27 '15 at 03:01

There's actually quite a bit of examples out there on this topic already I'd wager. It kind of depends on how exactly do you get your dates (which format etc..).

If you're not doing something overly out of this world and settle with the datetime module you can take a look at the example bellow. Fair warning from personal experience, locators and formatters are susceptible to weirdest bugs once the number of ticks reaches a very high number.

import matplotlib.pyplot as plt
import matplotlib as mpl
import numpy as np
from datetime import date, timedelta

#####################################
###  Creating data for plot, 
###  PLEASE include this in your next question.
#####################################
times = []
startdate = date(2014, 1, 1)
enddate = date(2016, 12, 22)
while startdate < enddate:
    times.append(startdate)
    startdate += timedelta(days=1) 

x = []
for i in range(0, len(times)):
    x.append(np.random.random(10))

First example, per months and days. The ticks for days are there, I just haven't shown them on the graph.

months = mpl.dates.MonthLocator()
days = mpl.dates.DayLocator(1) #find every 1st day of month
monthsFormatter = mpl.dates.DateFormatter('%b')

fig, ax = plt.subplots()

ax.xaxis.set_major_locator(months)
ax.xaxis.set_major_formatter(monthsFormatter)
ax.xaxis.set_minor_locator(days)

ax.plot(times, x)

plt.show()

Second example is something like this, per year basis with a more "detailed" formatter.

years = mpl.dates.YearLocator()
months = mpl.dates.MonthLocator()
yearsFormatter = mpl.dates.DateFormatter('%Y:%b')

fig, ax = plt.subplots()

ax.xaxis.set_major_locator(years)
ax.xaxis.set_major_formatter(yearsFormatter)
ax.xaxis.set_minor_locator(months)

ax.plot(times, x)

plt.show()

which looks something like:

how to plot data selectively in python

1 Answers1