Develop Reference

Plotting 3D scatter plot on top of map or image [duplicate]

This topic has been touched here, but no indications were given as to how to create a 3D plot and insert an image in the (x,y) plane, at a specified z height.
So to come up with a simple and reproducible case, let's say that I create a 3D plot like this with mplot3d:
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm
from matplotlib.ticker import LinearLocator, FormatStrFormatter
import matplotlib.pyplot as plt
import numpy as np
fig = plt.figure()
ax = fig.gca(projection='3d')
X = np.arange(-5, 5, 0.25)
Y = np.arange(-5, 5, 0.25)
X, Y = np.meshgrid(X, Y)
R = np.sqrt(X**2 + Y**2)
Z = np.sin(R)
surf = ax.plot_surface(X, Y, Z, rstride=1, cstride=1, cmap=cm.winter,
linewidth=0, antialiased=True)
ax.set_zlim(-1.01, 1.01)
ax.zaxis.set_major_locator(LinearLocator(10))
ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f'))
plt.show()
Visually we have:
At the level z=min(z)-1, where -1 is a visual offset to avoid overlapping, I want to insert an image representing the elements for which the curve shows a certain value. How to do it?
In this example I don't care about a perfect matching between the element and its value, so please feel free to upload any image you like. Also, is there a way of letting that image rotate, in case one is not happy with the matching?
EDIT
This is a visual example of something similar made for a 3D histogram. The grey shapes at the level z=0 are the elements for which the bars show a certain z value. Source.

Use plot_surface to draw image via facecolors argument.
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm
from matplotlib.ticker import LinearLocator, FormatStrFormatter
import matplotlib.pyplot as plt
import numpy as np
from matplotlib._png import read_png
from matplotlib.cbook import get_sample_data
fig = plt.figure()
ax = fig.gca(projection='3d')
X = np.arange(-5, 5, .25)
Y = np.arange(-5, 5, .25)
X, Y = np.meshgrid(X, Y)
R = np.sqrt(X**2 + Y**2)
Z = np.sin(R)
surf = ax.plot_surface(X, Y, Z, rstride=1, cstride=1, cmap=cm.winter,
linewidth=0, antialiased=True)
ax.set_zlim(-2.01, 1.01)
ax.zaxis.set_major_locator(LinearLocator(10))
ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f'))
fn = get_sample_data("./lena.png", asfileobj=False)
arr = read_png(fn)
# 10 is equal length of x and y axises of your surface
stepX, stepY = 10. / arr.shape[0], 10. / arr.shape[1]
X1 = np.arange(-5, 5, stepX)
Y1 = np.arange(-5, 5, stepY)
X1, Y1 = np.meshgrid(X1, Y1)
# stride args allows to determine image quality
# stride = 1 work slow
ax.plot_surface(X1, Y1, -2.01, rstride=1, cstride=1, facecolors=arr)
plt.show()
If you need to add values use PathPatch:
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm
from matplotlib.ticker import LinearLocator, FormatStrFormatter
import matplotlib.pyplot as plt
import numpy as np
from mpl_toolkits.mplot3d import Axes3D
import mpl_toolkits.mplot3d.art3d as art3d
from matplotlib.text import TextPath
from matplotlib.transforms import Affine2D
from matplotlib.patches import PathPatch
def text3d(ax, xyz, s, zdir="z", size=None, angle=0, usetex=False, **kwargs):
x, y, z = xyz
if zdir == "y":
xy1, z1 = (x, z), y
elif zdir == "y":
xy1, z1 = (y, z), x
else:
xy1, z1 = (x, y), z
text_path = TextPath((0, 0), s, size=size, usetex=usetex)
trans = Affine2D().rotate(angle).translate(xy1[0], xy1[1])
p1 = PathPatch(trans.transform_path(text_path), **kwargs)
ax.add_patch(p1)
art3d.pathpatch_2d_to_3d(p1, z=z1, zdir=zdir)
# main
fig = plt.figure()
ax = fig.gca(projection='3d')
X = np.arange(-5, 5, .25)
Y = np.arange(-5, 5, .25)
Xg, Yg = np.meshgrid(X, Y)
R = np.sqrt(Xg**2 + Yg**2)
Z = np.sin(R)
surf = ax.plot_surface(Xg, Yg, Z, rstride=1, cstride=1, cmap=cm.winter,
linewidth=0, antialiased=True)
ax.set_zlim(-2.01, 1.01)
ax.zaxis.set_major_locator(LinearLocator(10))
ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f'))
# add pathces with values
for i,x in enumerate(X[::4]):
for j,y in enumerate(Y[::4]):
text3d(ax, (x, y, -2.01), "{0:.1f}".format(Z[i][j]), zdir="z", size=.5, ec="none", fc="k")
plt.show()

How to plot horizontal stack of heatmaps or a stack of grid?

I want to plot a stack of heatmaps, contour, or grid computed over time. The plot should like this,
I have tried this:
from mpl_toolkits.mplot3d import Axes3D
import numpy as np
import matplotlib.pyplot as plt
fig = plt.figure()
ax = fig.gca(projection='3d')
x = np.linspace(0, 1, 100)
X, Z = np.meshgrid(x, x)
Y = np.sin(X)*np.sin(Z)
levels = np.linspace(-1, 1, 40)
ax.contourf(X, Y, Z, zdir='y')
ax.contourf(X, Y+3, Z, zdir='y')
ax.contourf(X, Y+7, Z, zdir='y')
ax.legend()
ax.view_init(15,155)
plt.show()
For one my plot looks ugly. It also does not look like what I want. I cannot make a grid there, and the 2d surfaces are tilted.
Any help is really appreciated! I am struggling with this.
Related stackoverflow:
[1] Python plot - stacked image slices
[2] Stack of 2D plot

How about making a series of 3d surface plots, with the data your wish to present in contour plotted as facecolor?
from mpl_toolkits.mplot3d import Axes3D
import matplotlib.pyplot as plt
from matplotlib import cm
from matplotlib.ticker import LinearLocator
import numpy as np
fig = plt.figure()
ax = fig.gca(projection='3d')
X = np.arange(-5, 5, 0.25)
Z = np.arange(-5, 5, 0.25)
X, Z = np.meshgrid(X, Z)
C = np.random.random(size=40*40*3).reshape((40, 40, 3))
ax.plot_surface(X, np.ones(shape=X.shape)-1, Z, facecolors=C, linewidth=0)
ax.plot_surface(X, np.ones(shape=X.shape), Z, facecolors=C, linewidth=0)
ax.plot_surface(X, np.ones(shape=X.shape)+1, Z, facecolors=C, linewidth=0)

Surface_plot: Add legend to facecolors

Let's say I'm plotting something with two different surface colors, as follows in colors red and blue. I want to add two items to the legend, with a custom string and the color for each color in my set.
How would I do that?
# generate the plot
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm
from matplotlib.ticker import LinearLocator, FormatStrFormatter
import matplotlib.pyplot as plt
import numpy as np
from numpy import random
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
X = np.arange(-5, 5, 0.25)
Y = np.arange(-5, 5, 0.25)
X, Y = np.meshgrid(X, Y)
R = np.sqrt(X**2 + Y**2)
Z = np.sin(R)
col1, col2 = cm.jet(np.array([0.1, 0.9]))
my_choice = random.choice([0, 1], size=X.shape)
my_color = my_choice[..., None] * col1[None, None, :] + (1 - my_choice)[..., None] * col2[None, None, :]
surf = ax.plot_surface(X, Y, Z, rstride=1, cstride=1, facecolors = my_color,
linewidth=0, antialiased=False)
ax.set_zlim(-1.01, 1.01)
# customstrings for legend:
myLegendLabels = {0: 'very red', 1: 'very blue'}

You can do that with "proxy artists":
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm
from matplotlib.ticker import LinearLocator, FormatStrFormatter
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches
import numpy as np
from numpy import random
# generate the plot
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
X = np.arange(-5, 5, 0.25)
Y = np.arange(-5, 5, 0.25)
X, Y = np.meshgrid(X, Y)
R = np.sqrt(X**2 + Y**2)
Z = np.sin(R)
col1, col2 = cm.jet(np.array([0.1, 0.9]))
my_choice = random.choice([0, 1], size=X.shape)
my_color = my_choice[..., None] * col1[None, None, :] + (1 - my_choice)[..., None] * col2[None, None, :]
surf = ax.plot_surface(X, Y, Z, rstride=1, cstride=1, facecolors = my_color,
linewidth=0, antialiased=False)
ax.set_zlim(-1.01, 1.01)
# Add legend with proxy artists
col1_patch = mpatches.Patch(color=col1, label='very blue')
col2_patch = mpatches.Patch(color=col2, label='very red')
plt.legend(handles=[col1_patch, col2_patch])
Result:

Set matplotlib view to be normal to the x-y plane in Python

This code found here is an example of a 3d surface plot:
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm
from matplotlib.ticker import LinearLocator, FormatStrFormatter
import matplotlib.pyplot as plt
import numpy as np
fig = plt.figure()
ax = fig.gca(projection='3d')
X = np.arange(-5, 5, 0.25)
Y = np.arange(-5, 5, 0.25)
X, Y = np.meshgrid(X, Y)
R = np.sqrt(X**2 + Y**2)
Z = np.sin(R)
surf = ax.plot_surface(X, Y, Z, rstride=1, cstride=1, cmap=cm.coolwarm,
linewidth=0, antialiased=False)
ax.set_zlim(-1.01, 1.01)
ax.zaxis.set_major_locator(LinearLocator(10))
ax.zaxis.set_major_formatter(FormatStrFormatter('%.02f'))
fig.colorbar(surf, shrink=0.5, aspect=5)
plt.show()
and yields
Is there a way to set the plot view so that it is perfectly normal to the x-y axis? This basically turns the 3-d plot into a 2-d one, where you can use the colourmap to judge the magnitude of the z-variable, rather than its displacement from the z=0 datum.

What you want is the ax.view_init function, with elev=90. See this answer
Edit:
after adding ax.view_init(azim=0, elev=90) to your script, I get this:

You need pcolor for that:
import matplotlib.pyplot as plt
import numpy as np
dx, dy = 0.25, 0.25
y, x = np.mgrid[slice(-5, 5 + dy, dy),
slice(-5, 5 + dx, dx)]
R = np.sqrt(x**2 + y**2)
z = np.sin(R)
z = z[:-1, :-1]
z_min, z_max = -np.abs(z).max(), np.abs(z).max()
plt.subplot()
plt.pcolor(x, y, z, cmap='RdBu', vmin=z_min, vmax=z_max)
plt.axis([x.min(), x.max(), y.min(), y.max()])
plt.colorbar()
plt.show()
Additional demos are here

matplotlib surface plot extends past axis limits

How do I make a nice paraboloid in Matplotlib that looks like
All I can get is this,
where the top is not "cut off". I've tried just dropping all values of the Z array outside of the radius of the parabola at the top, but that gives very jagged edges. Can someone help me?
Here is my code:
from matplotlib import *
from mpl_toolkits.mplot3d import Axes3D
import matplotlib.pyplot as plt
import numpy as np
from pylab import *
import math
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
X = np.arange(-5, 5, 0.1)
Y = np.arange(-5, 5, 0.1)
X, Y = np.meshgrid(X, Y)
Z = (X**2 + Y**2)
ax.set_zlim(-10, 20)
ax.plot_surface(X, Y, Z, alpha=0.9, rstride=4, cstride=4, linewidth=0.5, cmap=cm.summer)
plt.show()

For future reference, I had a thought to parametrize the surface in cylindrical coordinates, and it looks exactly how I want it:
from matplotlib import *
from mpl_toolkits.mplot3d import Axes3D
import matplotlib.pyplot as plt
import numpy as np
from pylab import *
import math
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
r = T = np.arange(0, 2*pi, 0.01)
r, T = np.meshgrid(r, T)
#Parametrise it
X = r*np.cos(T)
Y = r*np.sin(T)
Z = r**2
ax.plot_surface(X, Y, Z, alpha=0.9, rstride=10, cstride=10, linewidth=0.5, cmap=cm.summer)
plt.show()
I guess it makes sense: when working with a cylindrical object, use cylindrical coordinates!

Manual data clipping
One approach I've seen that works is to manually clip the data; e.g. your example would be updated to
from matplotlib import *
from mpl_toolkits.mplot3d import Axes3D
import matplotlib.pyplot as plt
import numpy as np
from pylab import *
import math
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
X = np.arange(-5, 5, 0.1)
Y = np.arange(-5, 5, 0.1)
X, Y = np.meshgrid(X, Y)
Z = (X**2 + Y**2)
ax.set_zlim(-10, 20)
for i in range(len(X)):
for j in range(len(Y)):
if (Z[j,i] < -10) or (Z[j,i] > 20):
Z[j,i] = NaN
ax.plot_surface(X, Y, Z, alpha=0.9, rstride=4, cstride=4, linewidth=0.5, cmap=cm.summer)
plt.show()
Note
This can be done concisely for this case using
Z[Z>20] = NaN
Resulting in