I have the following code for generating a time series plot
import numpy as np
fig = plt.figure()
ax = fig.add_subplot(111)
series = pd.Series([np.sin(ii*np.pi) for ii in range(30)],
index=pd.date_range(start='2019-01-01', end='2019-12-31',
periods=30))
series.plot(ax=ax)
I want to set an automatic limit for x and y, I tried using ax.margins() but it does not seem to work:
ax.margins(y=0.1, x=0.05)
# even with
# ax.margins(y=0.1, x=5)
What I am looking for is an automatic method like padding=0.1 (10% of whitespace around the graph)
Pandas and matplotlib seem to be confused rather often while collaborating when axes have dates. For some reason in this case ax.margins doesn't work as expected with the x-axis.
Here is a workaround which does seem to do the job, explicitely moving the xlims:
xmargins = 0.05
ymargins = 0.1
ax.margins(y=ymargins)
x0, x1 = plt.xlim()
plt.xlim(x0-xmargins*(x1-x0), x1+xmargins*(x1-x0))
Alternatively, you could work directly with matplotlib's plot, which does work as expected applying the margins to the date axis.
ax.plot(series.index, series)
ax.margins(y=0.1, x=0.05)
PS: This post talks about setting use_sticky_edges to False and calling autoscale_view after setting the margins, but also that doesn't seem to work here.
ax.use_sticky_edges = False
ax.autoscale_view(scaley=True, scalex=True)
You can use ax.set_xlim and ax.set_ylim to set the x and y limits of your plot respectively.
import numpy as np
fig = plt.figure()
ax = fig.add_subplot(111)
series = pd.Series([np.sin(ii*np.pi) for ii in range(30)],
index=pd.date_range(start='2019-01-01', end='2019-12-31',
periods=30))
# set xlim to be a between certain dates
ax.set_xlim((pd.to_datetime('2019-01-01'), pd.to_datetime('2019-01-31'))
# set ylim to be between certain values
ax.set_ylim((-0.5, 0.5))
series.plot(ax=ax)
Related
I am trying to add a small line outside of my axis range which I want to use as a highly customized legend at a later stage. However, using axes.hlines changes the xlim of my axis, even though I specify transform = axes.transAxes. The xlim appears to be set such that the coordinates of the hlines are included in the datacoordinate range. Only, that these coordinates are meant to be axes coordinates, not data coordinates.
Here comes a minimal working example:
import numpy as np
import matplotlib.pyplot as plt
x_data = np.random.rand(10)+10
y_data = np.random.rand(10)
fig, ax = plt.subplots()
ax.scatter(x_data,y_data)
ax.hlines(0.5,1.1,1.2, transform = ax.transAxes, clip_on = False)
results in xlims being changed by the ax.hlines command:
while with ax.hlines being commented out one gets:
I have to plot several curves with very high xtick density, say 1000 date strings. To prevent these tick labels overlapping each other I manually set them to be 60 dates apart. Code below:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
ts_index = pd.period_range(start="20060429", periods=1000).strftime("%Y%m%d")
fig = plt.figure(1)
ax = plt.subplot(1, 1, 1)
tick_spacing = 60
for i in range(5):
plt.plot(ts_index, 1 + i * 0.01 * np.arange(0, 1000), label="group %d"%i)
plt.legend(loc='best')
plt.title(r'net value curves')
xticks = ax.get_xticks()
xlabels = ax.get_xticklabels()
ax.set_xticks(xticks[::tick_spacing])
ax.set_xticklabels(xlabels[::tick_spacing])
plt.xticks(rotation="vertical")
plt.xlabel(r'date')
plt.ylabel('net value')
plt.grid(True)
plt.show()
fig.savefig(r".\net_value_curves.png", )
fig.clf()
I'm running this piece of code in PyCharm Community Edition 2017.2.2 with a Python 3.6 kernel. Now comes the funny thing: whenever I ran the code in the normal "run" mode (i.e. just hit the execution button and let the code run "freely" till interruption or termination), then the figure I got would always miss xticklabels:
However, if I ran the code in "debug" mode and ran it step by step then I would get an expected figure with complete xticklabels:
This is really weird. Anyway, I just hope to find a way that can ensure me getting the desired output (the second figure) in the normal "run" mode. How can I modify my current code to achieve this?
Thanks in advance!
Your x axis data are strings. Hence you will get one tick per data point. This is probably not what you want. Instead use the dates to plot. Because you are using pandas, this is easily converted,
dates = pd.to_datetime(ts_index, format="%Y%m%d")
You may then get rid of your manual xtick locating and formatting, because matplotlib will automatically choose some nice tick locations for you.
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
ts_index = pd.period_range(start="20060429", periods=1000).strftime("%Y%m%d")
dates = pd.to_datetime(ts_index, format="%Y%m%d")
fig, ax = plt.subplots()
for i in range(5):
plt.plot(dates, 1 + i * 0.01 * np.arange(0, 1000), label="group %d"%i)
plt.legend(loc='best')
plt.title(r'net value curves')
plt.xticks(rotation="vertical")
plt.xlabel(r'date')
plt.ylabel('net value')
plt.grid(True)
plt.show()
However in case you do want to have some manual control over the locations and formats you may use matplotlib.dates locators and formatters.
# tick every 3 months
plt.gca().xaxis.set_major_locator(mdates.MonthLocator((1,4,7,10)))
# format as "%Y%m%d"
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter("%Y%m%d"))
In general, the Axis object computes and places ticks using a Locator object. Locators and Formatters are meant to be easily replaceable, with appropriate methods of Axis. The default Locator does not seem to be doing the trick for you so you can replace it with anything you want using axes.xaxis.set_major_locator. This problem is not complicated enough to write your own, so I would suggest that MaxNLocator fits your needs fairly well. Your example seems to work well with nbins=16 (which is what you have in the picture, since there are 17 ticks.
You need to add an import:
from matplotlib.ticker import MaxNLocator
You need to replace the block
xticks = ax.get_xticks()
xlabels = ax.get_xticklabels()
ax.set_xticks(xticks[::tick_spacing])
ax.set_xticklabels(xlabels[::tick_spacing])
with
ax.xaxis.set_major_locator(MaxNLocator(nbins=16))
or just
ax.xaxis.set_major_locator(MaxNLocator(16))
You may want to play around with the other arguments (all of which have to be keywords, except nbins). Pay especial attention to integer.
Note that for the Locator and Formatter APIs we work with an Axis object, not Axes. Axes is the whole plot, while Axis is the thing with the spines on it. Axes usually contains two Axis objects and all the other stuff in your plot.
You can set the visibility of the xticks-labels to False
for label in plt.gca().xaxis.get_ticklabels()[::N]:
label.set_visible(False)
This will set every Nth label invisible.
I am trying to plot multiple lines in a 3D figure. Each line represents a month: I want them displayed parallel in the y-direction.
My plan was to loop over a set of Y values, but I cannot make this work properly, as using the ax.plot command (see working code below) produces a dozen lines all at the position of the final Y value. Confusingly, swapping ax.plot for ax.scatter does produce a set of parallel lines of data (albeit in the form of a set of dots; ax.view_init set to best display the parallel aspect of the result).
How can I use a produce a plot with multiple parallel lines?
My current workaround is to replace the loop with a dozen different arrays of Y values, and that can't be the right answer.
from mpl_toolkits.mplot3d.axes3d import Axes3D
import matplotlib.pyplot as plt
import numpy as np
# preamble
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
cs = ['r','g','b','y','r','g','b','y','r','g','b','y']
# x axis
X = np.arange(24)
# y axis
y = np.array([15,45,75,105,135,165,195,225,255,285,315,345])
Y = np.zeros(24)
# data - plotted against z axis
Z = np.random.rand(24)
# populate figure
for step in range(0,12):
Y[:] = y[step]
# ax.plot(X,Y,Z, color=cs[step])
ax.scatter(X,Y,Z, color=cs[step])
ax.set_xlabel('X')
ax.set_ylabel('Y')
ax.set_zlabel('Z')
# set initial view of plot
ax.view_init(elev=80., azim=345.)
plt.show()
I'm still learning python, so simple solutions (or, preferably, those with copious explanatory comments) are greatly appreciated.
Use
ax.plot(X, np.array(Y), Z, color=cs[step])
or
Y = [y[step]] * 24
This looks like a bug in mpl where we are not copying data when you hand it in so each line is sharing the same np.array object so when you update it all of your lines.
I am having an issue trying to get my date ticks rotated in matplotlib. A small sample program is below. If I try to rotate the ticks at the end, the ticks do not get rotated. If I try to rotate the ticks as shown under the comment 'crashes', then matplot lib crashes.
This only happens if the x-values are dates. If I replaces the variable dates with the variable t in the call to avail_plot, the xticks(rotation=70) call works just fine inside avail_plot.
Any ideas?
import numpy as np
import matplotlib.pyplot as plt
import datetime as dt
def avail_plot(ax, x, y, label, lcolor):
ax.plot(x,y,'b')
ax.set_ylabel(label, rotation='horizontal', color=lcolor)
ax.get_yaxis().set_ticks([])
#crashes
#plt.xticks(rotation=70)
ax2 = ax.twinx()
ax2.plot(x, [1 for a in y], 'b')
ax2.get_yaxis().set_ticks([])
ax2.set_ylabel('testing')
f, axs = plt.subplots(2, sharex=True, sharey=True)
t = np.arange(0.01, 5, 1)
s1 = np.exp(t)
start = dt.datetime.now()
dates=[]
for val in t:
next_val = start + dt.timedelta(0,val)
dates.append(next_val)
start = next_val
avail_plot(axs[0], dates, s1, 'testing', 'green')
avail_plot(axs[1], dates, s1, 'testing2', 'red')
plt.subplots_adjust(hspace=0, bottom=0.3)
plt.yticks([0.5,],("",""))
#doesn't crash, but does not rotate the xticks
#plt.xticks(rotation=70)
plt.show()
If you prefer a non-object-oriented approach, move plt.xticks(rotation=70) to right before the two avail_plot calls, eg
plt.xticks(rotation=70)
avail_plot(axs[0], dates, s1, 'testing', 'green')
avail_plot(axs[1], dates, s1, 'testing2', 'red')
This sets the rotation property before setting up the labels. Since you have two axes here, plt.xticks gets confused after you've made the two plots. At the point when plt.xticks doesn't do anything, plt.gca() does not give you the axes you want to modify, and so plt.xticks, which acts on the current axes, is not going to work.
For an object-oriented approach not using plt.xticks, you can use
plt.setp( axs[1].xaxis.get_majorticklabels(), rotation=70 )
after the two avail_plot calls. This sets the rotation on the correct axes specifically.
Solution works for matplotlib 2.1+
There exists an axes method tick_params that can change tick properties. It also exists as an axis method as set_tick_params
ax.tick_params(axis='x', rotation=45)
Or
ax.xaxis.set_tick_params(rotation=45)
As a side note, the current solution mixes the stateful interface (using pyplot) with the object-oriented interface by using the command plt.xticks(rotation=70). Since the code in the question uses the object-oriented approach, it's best to stick to that approach throughout. The solution does give a good explicit solution with plt.setp( axs[1].xaxis.get_majorticklabels(), rotation=70 )
An easy solution which avoids looping over the ticklabes is to just use
fig.autofmt_xdate()
This command automatically rotates the xaxis labels and adjusts their position. The default values are a rotation angle 30° and horizontal alignment "right". But they can be changed in the function call
fig.autofmt_xdate(bottom=0.2, rotation=30, ha='right')
The additional bottom argument is equivalent to setting plt.subplots_adjust(bottom=bottom), which allows to set the bottom axes padding to a larger value to host the rotated ticklabels.
So basically here you have all the settings you need to have a nice date axis in a single command.
A good example can be found on the matplotlib page.
Another way to applyhorizontalalignment and rotation to each tick label is doing a for loop over the tick labels you want to change:
import numpy as np
import matplotlib.pyplot as plt
import datetime as dt
now = dt.datetime.now()
hours = [now + dt.timedelta(minutes=x) for x in range(0,24*60,10)]
days = [now + dt.timedelta(days=x) for x in np.arange(0,30,1/4.)]
hours_value = np.random.random(len(hours))
days_value = np.random.random(len(days))
fig, axs = plt.subplots(2)
fig.subplots_adjust(hspace=0.75)
axs[0].plot(hours,hours_value)
axs[1].plot(days,days_value)
for label in axs[0].get_xmajorticklabels() + axs[1].get_xmajorticklabels():
label.set_rotation(30)
label.set_horizontalalignment("right")
And here is an example if you want to control the location of major and minor ticks:
import numpy as np
import matplotlib.pyplot as plt
import datetime as dt
fig, axs = plt.subplots(2)
fig.subplots_adjust(hspace=0.75)
now = dt.datetime.now()
hours = [now + dt.timedelta(minutes=x) for x in range(0,24*60,10)]
days = [now + dt.timedelta(days=x) for x in np.arange(0,30,1/4.)]
axs[0].plot(hours,np.random.random(len(hours)))
x_major_lct = mpl.dates.AutoDateLocator(minticks=2,maxticks=10, interval_multiples=True)
x_minor_lct = matplotlib.dates.HourLocator(byhour = range(0,25,1))
x_fmt = matplotlib.dates.AutoDateFormatter(x_major_lct)
axs[0].xaxis.set_major_locator(x_major_lct)
axs[0].xaxis.set_minor_locator(x_minor_lct)
axs[0].xaxis.set_major_formatter(x_fmt)
axs[0].set_xlabel("minor ticks set to every hour, major ticks start with 00:00")
axs[1].plot(days,np.random.random(len(days)))
x_major_lct = mpl.dates.AutoDateLocator(minticks=2,maxticks=10, interval_multiples=True)
x_minor_lct = matplotlib.dates.DayLocator(bymonthday = range(0,32,1))
x_fmt = matplotlib.dates.AutoDateFormatter(x_major_lct)
axs[1].xaxis.set_major_locator(x_major_lct)
axs[1].xaxis.set_minor_locator(x_minor_lct)
axs[1].xaxis.set_major_formatter(x_fmt)
axs[1].set_xlabel("minor ticks set to every day, major ticks show first day of month")
for label in axs[0].get_xmajorticklabels() + axs[1].get_xmajorticklabels():
label.set_rotation(30)
label.set_horizontalalignment("right")
Simply use
ax.set_xticklabels(label_list, rotation=45)
I am clearly late but there is an official example which uses
plt.setp(ax.get_xticklabels(), rotation=45, ha="right", rotation_mode="anchor")
to rotate the labels while keeping them correctly aligned with the ticks, which is both clean and easy.
Ref: https://matplotlib.org/stable/gallery/images_contours_and_fields/image_annotated_heatmap.html
I have the following Python code which I am using to plot a filled contour plot:
def plot_polar_contour(values, azimuths, zeniths):
theta = np.radians(azimuths)
zeniths = np.array(zeniths)
values = np.array(values)
values = values.reshape(len(azimuths), len(zeniths))
r, theta = np.meshgrid(zeniths, np.radians(azimuths))
fig, ax = subplots(subplot_kw=dict(projection='polar'))
ax.set_theta_zero_location("N")
ax.set_theta_direction(-1)
cax = ax.contourf(theta, r, values, 30)
autumn()
cb = fig.colorbar(cax)
cb.set_label("Pixel reflectance")
show()
This gives me a plot like:
However, when I add the line ax.plot(0, 30, 'p') just before show() I get the following:
It seems that just adding that one point (which is well within the original axis range) screws up the axis range on the radius axis.
Is this by design, or is this a bug? What would you suggest doing to fix it? Do I need to manually adjust the axis ranges, or is there a way to stop the extra plot command doing this?
If the axis auto-scaling mode isn't explicitly specified, plot will use "loose" autoscaling and contourf will use "tight" autoscaling.
The same things happens for non-polar axes. E.g.
import matplotlib.pyplot as plt
import numpy as np
plt.imshow(np.random.random((10,10)))
plt.plot([7], [7], 'ro')
plt.show()
You have a number of options.
Explicitly call ax.axis('image') or ax.axis('tight') at some point in the code.
Pass in scalex=False and scaley=False as keyword arguments to plot.
Manually set the axis limits.
The easiest and most readable is to just explicitly call ax.axis('tight'), i.m.o.