How to map multiple heatmap plots on one radar plot in python? - python

I have three data distributions:
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.colors import LogNorm
a = np.load('A.npy')
b = np.load('B.npy')
c = np.load('C.npy')
plt.figure(figsize=(12,4))
plt.subplot(131)
plt.hist2d(a,b,bins=300,norm=LogNorm())
plt.xlabel('A')
plt.ylabel('B')
plt.subplot(132)
plt.hist2d(a,c,bins=300,norm=LogNorm())
plt.xlabel('A')
plt.ylabel('C')
plt.subplot(133)
plt.hist2d(b,c,bins=300,norm=LogNorm())
plt.xlabel('B')
plt.ylabel('C')
plt.show()
And here is the result:
Now, I want to represent all three plots on a radar plot to look something like this:
Any ideas?

First, the easier part, the plotting: I've used 3 times the same random data, shrinked (0..2pi -> 0..2/3pi) and shifted (0, 2/3pi, 4/3pi) them to get 3 big pizza parts:
import numpy as np
import matplotlib.pyplot as plt
parts = 3
ax = plt.subplot(111, polar=True)
shrink = 1./parts
for i in range(3):
# beginning and end angle of this part
start = i * 2/parts * np.pi
end = (i + 1) * 2/parts * np.pi
# Generate random data:
N = 10000
r = .5 + np.random.normal(size=N, scale=.2)
theta = (np.pi / 2 + np.random.normal(size=N, scale=.1))
# shift the data counterclockwise so that it fills the n-th part
theta += i * 2.*np.pi / parts
# Histogramming
nr = 50
ntheta = 200
r_edges = np.linspace(0, 1, nr + 1)
theta_edges = np.linspace(start, end, ntheta + 1)
H, _, _ = np.histogram2d(r, theta, [r_edges, theta_edges])
# Plot
Theta, R = np.meshgrid(theta_edges, r_edges)
ax.pcolormesh(Theta, R, H)
plt.show()
Now the harder part: You will still have to convert your points into radial values, I'm not sure how you define this for your coordinates, as a point has 3 dimensions but you want to map to 2d. I hope this helps!
My code is based on this 2d heatmap.

Related

Colormap a 3D curve in matplotlib

I have 4 arrays x, y, z and T of length n and I want to plot a 3D curve using matplotlib. The (x, y, z) are the points positions and T is the value of each point (which is plotted as color), like the temperature of each point. How can I do it?
Example code:
import numpy as np
from matplotlib import pyplot as plt
fig = plt.figure()
ax = fig.gca(projection='3d')
n = 100
cmap = plt.get_cmap("bwr")
theta = np.linspace(-4 * np.pi, 4 * np.pi, n)
z = np.linspace(-2, 2, n)
r = z**2 + 1
x = r * np.sin(theta)
y = r * np.cos(theta)
T = (2*np.random.rand(n) - 1) # All the values are in [-1, 1]
What I found over the internet:
It's possible to use cmap with scatter like shown in the docs and in this stackoverflow question
ax = plt.gca()
ax.scatter(x, y, z, cmap=cmap, c=T)
The problem is that scatter is a set of points, not a curve.
In this stackoverflow question the solution was divide in n-1 intervals and each interval we use a different color like
t = (T - np.min(T))/(np.max(T)-np.min(T)) # Normalize
for i in range(n-1):
plt.plot(x[i:i+2], y[i:i+2], z[i:i+2], c=cmap(t[i])
The problem is that each segment has only one color, but it should be an gradient. The last value is not even used.
Useful links:
Matplotlib - Colormaps
Matplotlib - Tutorial 3D
This is a case where you probably need to use Line3DCollection. This is the recipe:
create segments from your array of coordinates.
create a Line3DCollection object.
add that collection to the axis.
set the axis limits.
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d.art3d import Line3DCollection
from matplotlib.cm import ScalarMappable
from matplotlib.colors import Normalize
def get_segments(x, y, z):
"""Convert lists of coordinates to a list of segments to be used
with Matplotlib's Line3DCollection.
"""
points = np.ma.array((x, y, z)).T.reshape(-1, 1, 3)
return np.ma.concatenate([points[:-1], points[1:]], axis=1)
fig = plt.figure()
ax = fig.add_subplot(projection='3d')
n = 100
cmap = plt.get_cmap("bwr")
theta = np.linspace(-4 * np.pi, 4 * np.pi, n)
z = np.linspace(-2, 2, n)
r = z**2 + 1
x = r * np.sin(theta)
y = r * np.cos(theta)
T = np.cos(theta)
segments = get_segments(x, y, z)
c = Line3DCollection(segments, cmap=cmap, array=T)
ax.add_collection(c)
fig.colorbar(c)
ax.set_xlim(x.min(), x.max())
ax.set_ylim(y.min(), y.max())
ax.set_zlim(z.min(), z.max())
plt.show()

Extend a 2D plot to 3D

I'm trying to show my 2D data on a 3D space.
Here is my code below:
import numpy as np
import matplotlib.pyplot as plt
i = 60
n = 1000
r = 3.8
eps = 0.7
y = np.ones((n, i))
# random numbers on the first row of array x
np.random.seed(1)
x = np.ones((n+1, i))
x[0, :] = np.random.random(i)
def logistic(r, x):
return r * x * (1 - x)
present_indi = np.arange(i)
next_indi = (present_indi + 1) % i
prev_indi = (present_indi - 1) % i
for n in range(1000):
y[n, :] = logistic(r, x[n, :])
x[n+1, :] = (1-eps)*y[n, present_indi] + 0.5*eps*(y[n, prev_indi] + y[n, next_indi])
#print(x)
# the above logic generates a 2D array 'x'. with i columns and n rows.
fig, ax = plt.subplots()
for i in range(60):
for n in range(1000):
if n>=900:
ax.plot(i,x[n,i],'*k',ms=0.9)
plt.xlabel('i')
plt.ylabel('x')
plt.title('test')
plt.show()
The above code perfectly displays i and x graph. I have plotted all the elements of 1st column of X, then all elements of second column, then the third and so on....., using the nested for loop logic (refer to the code)
Now what I need to do is, extend the plotting to 3D, i.e use Xaxis = i, Yaxis= n, Zaxis= array 'x'
I want to plot something like this:
I know I have to use mplot3D
But doing the following won't give me any result:
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
for i in range(60):
for n in range(1000):
if n>=900:
ax.plot_wireframe(i,n,x[n,i],rstride=1,cstride=1)
Plotting 3d images in matplotlib is a little tricky. Generally you plot whole surfaces at once instead of plotting one line at a time. You do so by passing three 2d arrays, one for each position dimension (x, y, z). But you can't just pass any old 2d arrays either; the points themselves have to be in a precise order!
Sometimes you can do something that just works, but I find it easier to explicitly parameterize plots using u and v dimensions. Here's what I was able to get working here:
# Abstract u and v parameters describing surface coordinates
u_plt = np.arange(x.shape[1])
v_plt = np.arange(x.shape[0])
# The outer products here produce 2d arrays. We multiply by
# ones in this case for an identity transformation, but in
# general, you could use any broadcasted operation on `u`
# and `v`.
x_plt = np.outer(np.ones(np.size(v_plt)), u_plt)
y_plt = np.outer(v_plt, np.ones(np.size(u_plt)))
# In this case, our `x` array gives the `z` values directly.
z_plt = x
fig = plt.figure(figsize=(16, 10))
ax = fig.add_subplot(111, projection='3d')
ax.set_zmargin(1) # Add a bit more space around the plot.
ax.plot_wireframe(x_plt, y_plt, z_plt,
rstride=1, cstride=1, # "Resolution" of the plot
color='blue', linewidth=1.0,
alpha=0.7, antialiased=True)
# Tilt the view to match the example.
ax.view_init(elev = 45, azim = -45)
plt.xlabel('i')
plt.ylabel('x')
plt.title('test')
plt.show()
And here's the resulting image. I had to reduce n to 80 to make this comprehensible at all, and I have no idea what I am looking at, so I am not sure it's correct. But I think it looks broadly similar to the example you gave.
Just to illustrate the power of this approach, here's a nautilus shell. It uses a two-stage parameterization, which could be compressed, but which I find conceptually clearer:
n_ticks = 100
# Abstract u and v parameters describing surface coordinates
u_plt = np.arange(n_ticks // 2) * 2
v_plt = np.arange(n_ticks)
# theta is the angle along the leading edge of the shell
# phi is the angle along the spiral of the shell
# r is the distance of the edge from the origin
theta_plt = np.pi * ((u_plt / n_ticks) * 0.99 + 0.005)
phi_plt = np.pi * v_plt / (n_ticks / 5)
r_plt = v_plt / (n_ticks / 5)
# These formulas are based on the formulas for rendering
# a sphere parameterized by theta and phi. The only difference
# is that r is variable here too.
x_plt = r_plt[:, None] * np.cos(phi_plt[:, None]) * np.sin(theta_plt[None, :])
y_plt = r_plt[:, None] * np.sin(phi_plt[:, None]) * np.sin(theta_plt[None, :])
z_plt = r_plt[:, None] * \
(np.ones(np.shape(phi_plt[:, None])) * np.cos(theta_plt[None, :]))
# This varies the color along phi
colors = cm.inferno(1 - (v_plt[:, None] / max(v_plt))) * \
np.ones(np.shape(u_plt[None, :, None]))
fig = plt.figure(figsize=(8, 8))
ax = fig.add_subplot(111, projection='3d')
ax.set_zmargin(1)
ax.plot_surface(x_plt, y_plt, z_plt,
rstride=1, cstride=1,
facecolors=colors, linewidth=1.0,
alpha=0.3, antialiased=True)
ax.view_init(elev = 45, azim = -45)
plt.show()

Aligning maps made using basemap

Is there a way to align python basemaps like this figure below?
Here's some sample basemap code to produce a map:
from mpl_toolkits.basemap import Basemap
import matplotlib.pyplot as plt
fig = plt.figure(figsize=(8, 4.5))
plt.subplots_adjust(left=0.02, right=0.98, top=0.98, bottom=0.00)
m = Basemap(projection='robin',lon_0=0,resolution='c')
m.fillcontinents(color='gray',lake_color='white')
m.drawcoastlines()
plt.savefig('world.png',dpi=75)
I am not an expert with Matplotlib, but I found a way to get a similar result by using the data files included in the source folder of basemap. They can be combined into a meshgrid to plot some data, in the example below we plot the altitude at every point.
One of the tricks I used is to set matplotlib to an orthogonal projection so that there is no distortion in the vertical spacing of the maps.
I have put the parameters at the beginning of the code as you may find it useful to adjust.
One thing I couldn't get my head around is the shadow under the maps.
from mpl_toolkits.mplot3d import proj3d
from mpl_toolkits.basemap import Basemap
from mpl_toolkits.axes_grid1.inset_locator import inset_axes
import numpy as np
import matplotlib.pyplot as plt
# Parameters
n_maps = 5 # Number of maps
z_spacing = 4. # Spacing of maps along z
z_reduction = 1E-8 # Reduction factor for Z data, makes the map look flat
view_angles = (14., -100.) # Set view port angles
colbar_bottom = 0.2 # Space at the bottom of colorbar column
colbar_spacing = .132 # Space between colorbars
colbar_height = 0.1 # Height of colorbars
# Set orthogonal projection
def orthogonal_proj(zfront, zback):
a = (zfront+zback)/(zfront-zback)
b = -2*(zfront*zback)/(zfront-zback)
return np.array([[1,0,0,0],
[0,1,0,0],
[0,0,a,b],
[0,0,-0.0001,zback]])
proj3d.persp_transformation = orthogonal_proj
fig = plt.figure(figsize=[30, 10*n_maps])
ax = fig.gca(projection='3d')
etopo = np.loadtxt('etopo20data.gz')
lons = np.loadtxt('etopo20lons.gz')
lats = np.loadtxt('etopo20lats.gz')
# Create Basemap instance for Robinson projection.
m = Basemap(projection='robin', lon_0=0.5*(lons[0]+lons[-1]))
# Compute map projection coordinates for lat/lon grid.
X, Y = m(*np.meshgrid(lons,lats))
# Exclude the oceans
Z = etopo.clip(-1)
# Set the colormap
cmap = plt.cm.get_cmap("terrain")
cmap.set_under("grey")
for i in range(n_maps):
c = ax.contourf(X, Y, z_spacing*i + z_reduction*Z, 30, cmap=cmap, vmin=z_spacing*i, extend='neither')
cax = inset_axes(ax,
width="5%",
height="100%",
loc=3,
bbox_to_anchor=(.85, colbar_spacing*i+colbar_bottom, .2, colbar_height),
bbox_transform=ax.transAxes,
borderpad=0
)
cb = fig.colorbar(c, cax=cax)
cb.set_label("Altitude")
# Reset the ticks of the color bar to match initial data
cb.set_ticks([z_spacing * i + j/10. * z_reduction * Z.max() for j in range(11)])
cb.set_ticklabels([str(int(j/10. * Z.max())) for j in range(11)])
ax.set_axis_off()
ax.view_init(*view_angles)
ax.set_xlim3d(X.min(), X.max())
ax.set_ylim3d(Y.min(), Y.max())
ax.set_zlim3d(-1E-2, (n_maps-1)*z_spacing)
plt.savefig('world.png',dpi=75)
Edit:
If you want shadows and don't mind the extra compute time you can change the beginning of the for loop with something along the lines of:
shadow_Z = np.empty(Z.shape)
for i in range(n_maps):
c = ax.contourf(X, Y, z_spacing*i + z_reduction*Z, 30, cmap=cmaps[i], vmin=z_spacing*i, extend='neither')
for j in range(10):
shadow_Z.fill(z_spacing*i - 1E-2 * j)
s = ax.contourf((X - X.mean()) * (1 + 8E-3 * j) + X.mean() + 2E5,
(Y - Y.mean()) * (1 + 8E-3 * j) + Y.mean() - 2E5,
shadow_Z, colors='black', alpha=0.1 - j * 1E-2)

Elevation distortion on sphere-projected image in Python

I'm trying to take two rectangular images, one of visible surface features and one representing elevation, and map them onto a 3D sphere. I know how to map features onto a sphere with Cartopy, and I know how to make relief surface maps, but I can't find a simple way to combine them to have exaggerated elevation on a spherical projection. For an example, here's it done in MATLAB:
Does anybody know if there's a simple way to do this in Python?
My solution does not meet all of your requirements. But it could be a good starter, to begin with.
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
import numpy as np
from matplotlib.cbook import get_sample_data
from matplotlib._png import read_png
# Use world image with shape (360 rows, 720 columns)
pngfile = 'temperature_15-115.png'
fn = get_sample_data(pngfile, asfileobj=False)
img = read_png(fn) # get array of color
# Some needed functions / constant
r = 5
pi = np.pi
cos = np.cos
sin = np.sin
sqrt = np.sqrt
# Prep values to match the image shape (360 rows, 720 columns)
phi, theta = np.mgrid[0:pi:360j, 0:2*pi:720j]
# Parametric eq for a distorted globe (for demo purposes)
x = r * sin(phi) * cos(theta)
y = r * sin(phi) * sin(theta)
z = r * cos(phi) + 0.5* sin(sqrt(x**2 + y**2)) * cos(2*theta)
fig = plt.figure()
fig.set_size_inches(9, 9)
ax = fig.add_subplot(111, projection='3d', label='axes1')
# Drape the image (img) on the globe's surface
sp = ax.plot_surface(x, y, z, \
rstride=2, cstride=2, \
facecolors=img)
ax.set_aspect(1)
plt.show()
The resulting image:

Vertically fill 3d matplotlib plot

I have a 3d plot made using matplotlib. I now want to fill the vertical space between the drawn line and the x,y axis to highlight the height of the line on the z axis. On a 2d plot this would be done with fill_between but there does not seem to be anything similar for a 3d plot. Can anyone help?
here is my current code
from stravalib import Client
import matplotlib as mpl
import numpy as np
import matplotlib.pyplot as plt
... code to get the data ....
mpl.rcParams['legend.fontsize'] = 10
fig = plt.figure()
ax = fig.gca(projection='3d')
zi = alt
x = df['x'].tolist()
y = df['y'].tolist()
ax.plot(x, y, zi, label='line')
ax.legend()
plt.show()
and the current plot
just to be clear I want a vertical fill to the x,y axis intersection NOT this...
You're right. It seems that there is no equivalent in 3D plot for the 2D plot function fill_between. The solution I propose is to convert your data in 3D polygons. Here is the corresponding code:
import math as mt
import matplotlib.pyplot as pl
import numpy as np
import random as rd
from mpl_toolkits.mplot3d import Axes3D
from mpl_toolkits.mplot3d.art3d import Poly3DCollection
# Parameter (reference height)
h = 0.0
# Code to generate the data
n = 200
alpha = 0.75 * mt.pi
theta = [alpha + 2.0 * mt.pi * (float(k) / float(n)) for k in range(0, n + 1)]
xs = [1.0 * mt.cos(k) for k in theta]
ys = [1.0 * mt.sin(k) for k in theta]
zs = [abs(k - alpha - mt.pi) * rd.random() for k in theta]
# Code to convert data in 3D polygons
v = []
for k in range(0, len(xs) - 1):
x = [xs[k], xs[k+1], xs[k+1], xs[k]]
y = [ys[k], ys[k+1], ys[k+1], ys[k]]
z = [zs[k], zs[k+1], h, h]
#list is necessary in python 3/remove for python 2
v.append(list(zip(x, y, z)))
poly3dCollection = Poly3DCollection(v)
# Code to plot the 3D polygons
fig = pl.figure()
ax = Axes3D(fig)
ax.add_collection3d(poly3dCollection)
ax.set_xlim([min(xs), max(xs)])
ax.set_ylim([min(ys), max(ys)])
ax.set_zlim([min(zs), max(zs)])
ax.set_xlabel("x")
ax.set_ylabel("y")
ax.set_zlabel("z")
pl.show()
It produces the following figure:
I hope this will help you.

Categories