Develop Reference

Multiple 2D histogram on same plot

I have this script to extract data from an image and roi. I have everything working perfectly except the end when I output the graphs. Basically I'm having trouble with the windowing of both histograms. It doesn't matter if I change the gridsize, mincount, figure size, or x and y limits one of the histograms will always be slightly stretched. When I plot them individually they aren't stretched. Is there a way to make the hexagons on the same plot a consistent "non-stretched" shape?
Down below is my graph and plotting methods. (I left out my data extraction methods because it was quite specialized).
plt.ion()
plt.figure(figsize=(16,8))
plt.title('2D Histogram of Entorhinal Cortex ROIs')
plt.xlabel(x_inputs)
plt.ylabel(y_inputs)
colors = ['Reds','Blues']
x = []
y= []
#image extraction code
hist1 = plt.hexbin(x[0],y[0], gridsize=100,cmap='Reds',mincnt=10, alpha=0.35)
hist2 = plt.hexbin(x[1],y[1], gridsize=100,cmap='Blues',mincnt=10, alpha=0.35)
plt.colorbar(hist1, orientation="vertical")
plt.colorbar(hist2, orientation="vertical")
plt.ioff()
plt.show()
enter image description here

This issue can be solved by setting limits for the bins with the extent parameter. This can be done automatically by computing the minimum and maximum x and y values across all the data being plotted. In cases where gridsize is small (e.g. 10), this approach may result in some of the bins being partially outside of the plot limits. If so, setting a margin with plt.margins can help display all the bins within the plot.
import numpy as np # v 1.20.2
import matplotlib.pyplot as plt # v 3.3.4
# Create a random dataset
rng = np.random.default_rng(seed=123) # random number generator
size = 10000
x1 = rng.normal(loc=5, scale=10, size=size)
y1 = rng.normal(loc=5, scale=2, size=size)
x2 = rng.normal(loc=-30, scale=5, size=size)
y2 = rng.normal(loc=-20, scale=5, size=size)
# Define hexbin grid extent
xmin = min(*x1, *x2)
xmax = max(*x1, *x2)
ymin = min(*y1, *y2)
ymax = max(*y1, *y2)
ext = (xmin, xmax, ymin, ymax)
# Draw figure with colorbars
plt.figure(figsize=(10, 6))
hist1 = plt.hexbin(x1, y1, gridsize=30, cmap='Reds', mincnt=10, alpha=0.3, extent=ext)
hist2 = plt.hexbin(x2, y2, gridsize=30, cmap='Blues', mincnt=10, alpha=0.3, extent=ext)
plt.colorbar(hist1, orientation='vertical')
plt.colorbar(hist2, orientation='vertical')
# plt.margins(0.1) # Uncomment this if hex bins are partially outside of plot limits
plt.show()

Python pcolormesh with separate alpha value for each bin

Lets say I have the following dataset:
import numpy as np
import matplotlib.pyplot as plt
x_bins = np.arange(10)
y_bins = np.arange(10)
z = np.random.random((9,9))
I can easily plot this data with
plt.pcolormesh(x_bins, y_bins, z, cmap = 'viridis)
However, let's say I now add some alpha value for each point:
a = np.random.random((9,9))
How can I change the alpha value of each box in the pcolormesh plot to match the corresponding value in array "a"?

The mesh created by pcolormesh can only have one alpha for the complete mesh. To set an individual alpha for each cell, the cells need to be created one by one as rectangles.
The code below shows the pcolormesh without alpha at the left, and the mesh of rectangles with alpha at the right. Note that on the spots where the rectangles touch, the semi-transparency causes some unequal overlap. This can be mitigated by not drawing the cell edge (edgecolor='none'), or by longer black lines to separate the cells.
The code below changes the x dimension so easier verify that x and y aren't mixed up. relim and autoscale are needed because with matplotlib's default behavior the x and y limits aren't changed by adding patches.
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.patches import Rectangle, Patch
x_bins = np.arange(12)
y_bins = np.arange(10)
z = np.random.random((9, 11))
a = np.random.random((9, 11))
cmap = plt.get_cmap('inferno')
norm = plt.Normalize(z.min(), z.max())
fig, (ax1, ax2) = plt.subplots(ncols=2)
ax1.pcolormesh(x_bins, y_bins, z, cmap=cmap, norm=norm)
for i in range(len(x_bins) - 1):
for j in range(len(y_bins) - 1):
rect = Rectangle((x_bins[i], y_bins[j]), x_bins[i + 1] - x_bins[i], y_bins[j + 1] - y_bins[j],
facecolor=cmap(norm(z[j, i])), alpha=a[j, i], edgecolor='none')
ax2.add_patch(rect)
# ax2.vlines(x_bins, y_bins.min(), y_bins.max(), edgecolor='black')
# ax2.hlines(y_bins, x_bins.min(), x_bins.max(), edgecolor='black')
ax2.relim()
ax2.autoscale(enable=True, tight=True)
plt.show()

plot with polycollection disappears when polygons get too small

I'm using a PolyCollection to plot data of various sizes. Sometimes the polygons are very small. If they are too small, they don't get plotted at all. I would expect the outline at least to show up so you'd have an idea that some data is there. Is there a a setting to control this?
Here's some code to reproduce the problem, as well as the output image:
import matplotlib.pyplot as plt
from matplotlib.collections import PolyCollection
from matplotlib import colors
fig = plt.figure()
ax = fig.add_subplot(111)
verts = []
edge_col = colors.colorConverter.to_rgb('lime')
face_col = [(2.0 + val) / 3.0 for val in edge_col] # a little lighter
for i in range(10):
w = 0.5 * 10**(-i)
xs = [i - w, i - w, i + w, i - w]
ys = [-w, w, 0, -w]
verts.append(list(zip(xs, ys)))
ax.set_xlim(-1, 11)
ax.set_ylim(-2, 2)
ax.add_collection(PolyCollection(verts, lw=3, alpha=0.5, edgecolor=edge_col, facecolor=face_col))
plt.savefig('out.png')
Notice that only six polygons are visible, whereas there should be ten.
Edit: I understand I could zoom in to see the others, but I was hoping to see a dot or the outline or something without doing this.
Edit 2: Here's a way to get the desired effect, by plotting the faces using a PolyCollection and then the edges using a series of Line2D plots with markers, based on Patol75's answer. My application is a matplotlib animation with lots of polygons, so I'd prefer to avoid Line2D for efficiency, and it would be cleaner if I didn't need to plot things twice, so I'm still hoping for a better answer.
ax.add_collection(PolyCollection(verts, lw=3, alpha=0.5, edgecolor=None, facecolor=face_col, zorder=1))
for pts in verts:
ax.add_line(Line2D([pt[0] for pt in pts], [pt[1] for pt in pts], lw=3, alpha=0.5, color=edge_col,
marker='.', ms=1, mec=edge_col, solid_capstyle='projecting', zorder=2))

Zooming in your plotting window, you would notice that your two remaining polygons are being plotted. They are just too small for you to see them. One way to be convinced of this is to replace
ax.set_xlim(-1, 6)
ax.set_ylim(-2, 2)
by
ax.set_xlim(1e-1, 1e1)
ax.set_ylim(1e-5, 1e0)
ax.set_xscale('log')
ax.set_yscale('log')
ax.set_aspect('equal')
Your five polygons are now visible, but on the downside the log scale restrains you to the positive side of the axes.
Now to propose an answer to your problem. If you keep a linear axis, as your polygons sizes span multiple orders of magnitude, you will not be able to see them all. What you can do is add an artist on your plot which specifies their location. This can be done with a marker, an arrow, etc... If we take the example of a marker, as you said, we only want to see this marker if we cannot see the polygon. The keyword zorder in the call to plot() allows to specify which artist should have the display priority on the figure. Please consider the example below.
import matplotlib.pyplot as plt
from matplotlib.collections import PolyCollection
fig = plt.figure()
ax = fig.add_subplot(111)
verts = []
for i in range(5):
w = 0.5 * 10**(-i)
xs = [i - w, i - w, i + w, i + w, i - w]
ys = [-w, w, w, -w, -w]
ax.plot((xs[2] + xs[1]) / 2, (ys[1] + ys[0]) / 2, linestyle='none',
marker='o', color='xkcd:crimson', markersize=1, zorder=-1)
verts.append(list(zip(xs, ys)))
ax.set_xlim(-1, 6)
ax.set_ylim(-2, 2)
poly = PolyCollection(verts, lw=5, edgecolor='black', facecolor='gray')
ax.add_collection(poly)
plt.show()
which produces
You would notice that if you zoom on the last two dots in the matplotlib figure, you actually do not see the markers, but rather the polygons.

You may introduce some minimal unit minw, which is the smallest size a shape can have.
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.collections import PolyCollection
from matplotlib import colors
fig = plt.figure()
ax = fig.add_subplot(111)
verts = []
edge_col = colors.colorConverter.to_rgb('lime')
face_col = [(2.0 + val) / 3.0 for val in edge_col] # a little lighter
ax.set_xlim(-1, 11)
ax.set_ylim(-2, 2)
u = np.diff(np.array([ax.get_xlim(), ax.get_ylim()]), axis=1).min()
px = np.max(fig.get_size_inches())*fig.dpi
minw = u/px/2
for i in range(10):
w = 0.5 * 10**(-i)
if w < minw:
w = minw
xs = [i - w, i - w, i + w, i - w]
ys = [-w, w, 0, -w]
verts.append(list(zip(xs, ys)))
ax.add_collection(PolyCollection(verts, lw=3, alpha=0.5, edgecolor=edge_col, facecolor=face_col))
plt.savefig('out.png')
plt.show()

Set a colormap under a graph

I know this is well documented, but I'm struggling to implement this in my code.
I would like to shade the area under my graph with a colormap. Is it possible to have a colour, i.e. red from any points over 30, and a gradient up until that point?
I am using the method fill_between, but I'm happy to change this if there is a better way to do it.
def plot(sd_values):
plt.figure()
sd_values=np.array(sd_values)
x=np.arange(len(sd_values))
plt.plot(x,sd_values, linewidth=1)
plt.fill_between(x,sd_values, cmap=plt.cm.jet)
plt.show()
This is the result at the moment. I have tried axvspan, but this doesnt have cmap as an option. Why does the below graph not show a colormap?

I'm not sure if the cmap argument should be part of the fill_between plotting command. In your case probably want to use the fill() command btw.
These fill commands create polygons or polygon collections. A polygon collection can take a cmap but with fill there is no way of providing the data on which it should be colored.
What's (for as far as i know) certainly not possible is to fill a single polygon with a gradient as you wish.
The next best thing is to fake it. You can plot a shaded image and clip it based on the created polygon.
# create some sample data
x = np.linspace(0, 1)
y = np.sin(4 * np.pi * x) * np.exp(-5 * x) * 120
fig, ax = plt.subplots()
# plot only the outline of the polygon, and capture the result
poly, = ax.fill(x, y, facecolor='none')
# get the extent of the axes
xmin, xmax = ax.get_xlim()
ymin, ymax = ax.get_ylim()
# create a dummy image
img_data = np.arange(ymin,ymax,(ymax-ymin)/100.)
img_data = img_data.reshape(img_data.size,1)
# plot and clip the image
im = ax.imshow(img_data, aspect='auto', origin='lower', cmap=plt.cm.Reds_r, extent=[xmin,xmax,ymin,ymax], vmin=y.min(), vmax=30.)
im.set_clip_path(poly)
The image is given an extent which basically stretches it over the entire axes. Then the clip_path makes it only showup where the fill polygon is drawn.

I think all you need is to do the plot of the data one at a time, like:
import numpy
import matplotlib.pyplot as plt
import matplotlib.cm as cm
import matplotlib.colors as colors
# Create fake data
x = numpy.linspace(0,4)
y = numpy.exp(x)
# Now plot one by one
bar_width = x[1] - x[0] # assuming x is linealy spaced
for pointx, pointy in zip(x,y):
current_color = cm.jet( min(pointy/30, 30)) # maximum of 30
plt.bar(pointx, pointy, bar_width, color = current_color)
plt.show()
Resulting in:

Generate a heatmap using a scatter data set

I have a set of X,Y data points (about 10k) that are easy to plot as a scatter plot but that I would like to represent as a heatmap.
I looked through the examples in Matplotlib and they all seem to already start with heatmap cell values to generate the image.
Is there a method that converts a bunch of x, y, all different, to a heatmap (where zones with higher frequency of x, y would be "warmer")?

If you don't want hexagons, you can use numpy's histogram2d function:
import numpy as np
import numpy.random
import matplotlib.pyplot as plt
# Generate some test data
x = np.random.randn(8873)
y = np.random.randn(8873)
heatmap, xedges, yedges = np.histogram2d(x, y, bins=50)
extent = [xedges[0], xedges[-1], yedges[0], yedges[-1]]
plt.clf()
plt.imshow(heatmap.T, extent=extent, origin='lower')
plt.show()
This makes a 50x50 heatmap. If you want, say, 512x384, you can put bins=(512, 384) in the call to histogram2d.
Example:

In Matplotlib lexicon, i think you want a hexbin plot.
If you're not familiar with this type of plot, it's just a bivariate histogram in which the xy-plane is tessellated by a regular grid of hexagons.
So from a histogram, you can just count the number of points falling in each hexagon, discretiize the plotting region as a set of windows, assign each point to one of these windows; finally, map the windows onto a color array, and you've got a hexbin diagram.
Though less commonly used than e.g., circles, or squares, that hexagons are a better choice for the geometry of the binning container is intuitive:
hexagons have nearest-neighbor symmetry (e.g., square bins don't,
e.g., the distance from a point on a square's border to a point
inside that square is not everywhere equal) and
hexagon is the highest n-polygon that gives regular plane
tessellation (i.e., you can safely re-model your kitchen floor with hexagonal-shaped tiles because you won't have any void space between the tiles when you are finished--not true for all other higher-n, n >= 7, polygons).
(Matplotlib uses the term hexbin plot; so do (AFAIK) all of the plotting libraries for R; still i don't know if this is the generally accepted term for plots of this type, though i suspect it's likely given that hexbin is short for hexagonal binning, which is describes the essential step in preparing the data for display.)
from matplotlib import pyplot as PLT
from matplotlib import cm as CM
from matplotlib import mlab as ML
import numpy as NP
n = 1e5
x = y = NP.linspace(-5, 5, 100)
X, Y = NP.meshgrid(x, y)
Z1 = ML.bivariate_normal(X, Y, 2, 2, 0, 0)
Z2 = ML.bivariate_normal(X, Y, 4, 1, 1, 1)
ZD = Z2 - Z1
x = X.ravel()
y = Y.ravel()
z = ZD.ravel()
gridsize=30
PLT.subplot(111)
# if 'bins=None', then color of each hexagon corresponds directly to its count
# 'C' is optional--it maps values to x-y coordinates; if 'C' is None (default) then
# the result is a pure 2D histogram
PLT.hexbin(x, y, C=z, gridsize=gridsize, cmap=CM.jet, bins=None)
PLT.axis([x.min(), x.max(), y.min(), y.max()])
cb = PLT.colorbar()
cb.set_label('mean value')
PLT.show()

Edit: For a better approximation of Alejandro's answer, see below.
I know this is an old question, but wanted to add something to Alejandro's anwser: If you want a nice smoothed image without using py-sphviewer you can instead use np.histogram2d and apply a gaussian filter (from scipy.ndimage.filters) to the heatmap:
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.cm as cm
from scipy.ndimage.filters import gaussian_filter
def myplot(x, y, s, bins=1000):
heatmap, xedges, yedges = np.histogram2d(x, y, bins=bins)
heatmap = gaussian_filter(heatmap, sigma=s)
extent = [xedges[0], xedges[-1], yedges[0], yedges[-1]]
return heatmap.T, extent
fig, axs = plt.subplots(2, 2)
# Generate some test data
x = np.random.randn(1000)
y = np.random.randn(1000)
sigmas = [0, 16, 32, 64]
for ax, s in zip(axs.flatten(), sigmas):
if s == 0:
ax.plot(x, y, 'k.', markersize=5)
ax.set_title("Scatter plot")
else:
img, extent = myplot(x, y, s)
ax.imshow(img, extent=extent, origin='lower', cmap=cm.jet)
ax.set_title("Smoothing with $\sigma$ = %d" % s)
plt.show()
Produces:
The scatter plot and s=16 plotted on top of eachother for Agape Gal'lo (click for better view):
One difference I noticed with my gaussian filter approach and Alejandro's approach was that his method shows local structures much better than mine. Therefore I implemented a simple nearest neighbour method at pixel level. This method calculates for each pixel the inverse sum of the distances of the n closest points in the data. This method is at a high resolution pretty computationally expensive and I think there's a quicker way, so let me know if you have any improvements.
Update: As I suspected, there's a much faster method using Scipy's scipy.cKDTree. See Gabriel's answer for the implementation.
Anyway, here's my code:
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.cm as cm
def data_coord2view_coord(p, vlen, pmin, pmax):
dp = pmax - pmin
dv = (p - pmin) / dp * vlen
return dv
def nearest_neighbours(xs, ys, reso, n_neighbours):
im = np.zeros([reso, reso])
extent = [np.min(xs), np.max(xs), np.min(ys), np.max(ys)]
xv = data_coord2view_coord(xs, reso, extent[0], extent[1])
yv = data_coord2view_coord(ys, reso, extent[2], extent[3])
for x in range(reso):
for y in range(reso):
xp = (xv - x)
yp = (yv - y)
d = np.sqrt(xp**2 + yp**2)
im[y][x] = 1 / np.sum(d[np.argpartition(d.ravel(), n_neighbours)[:n_neighbours]])
return im, extent
n = 1000
xs = np.random.randn(n)
ys = np.random.randn(n)
resolution = 250
fig, axes = plt.subplots(2, 2)
for ax, neighbours in zip(axes.flatten(), [0, 16, 32, 64]):
if neighbours == 0:
ax.plot(xs, ys, 'k.', markersize=2)
ax.set_aspect('equal')
ax.set_title("Scatter Plot")
else:
im, extent = nearest_neighbours(xs, ys, resolution, neighbours)
ax.imshow(im, origin='lower', extent=extent, cmap=cm.jet)
ax.set_title("Smoothing over %d neighbours" % neighbours)
ax.set_xlim(extent[0], extent[1])
ax.set_ylim(extent[2], extent[3])
plt.show()
Result:

Instead of using np.hist2d, which in general produces quite ugly histograms, I would like to recycle py-sphviewer, a python package for rendering particle simulations using an adaptive smoothing kernel and that can be easily installed from pip (see webpage documentation). Consider the following code, which is based on the example:
import numpy as np
import numpy.random
import matplotlib.pyplot as plt
import sphviewer as sph
def myplot(x, y, nb=32, xsize=500, ysize=500):
xmin = np.min(x)
xmax = np.max(x)
ymin = np.min(y)
ymax = np.max(y)
x0 = (xmin+xmax)/2.
y0 = (ymin+ymax)/2.
pos = np.zeros([len(x),3])
pos[:,0] = x
pos[:,1] = y
w = np.ones(len(x))
P = sph.Particles(pos, w, nb=nb)
S = sph.Scene(P)
S.update_camera(r='infinity', x=x0, y=y0, z=0,
xsize=xsize, ysize=ysize)
R = sph.Render(S)
R.set_logscale()
img = R.get_image()
extent = R.get_extent()
for i, j in zip(xrange(4), [x0,x0,y0,y0]):
extent[i] += j
print extent
return img, extent
fig = plt.figure(1, figsize=(10,10))
ax1 = fig.add_subplot(221)
ax2 = fig.add_subplot(222)
ax3 = fig.add_subplot(223)
ax4 = fig.add_subplot(224)
# Generate some test data
x = np.random.randn(1000)
y = np.random.randn(1000)
#Plotting a regular scatter plot
ax1.plot(x,y,'k.', markersize=5)
ax1.set_xlim(-3,3)
ax1.set_ylim(-3,3)
heatmap_16, extent_16 = myplot(x,y, nb=16)
heatmap_32, extent_32 = myplot(x,y, nb=32)
heatmap_64, extent_64 = myplot(x,y, nb=64)
ax2.imshow(heatmap_16, extent=extent_16, origin='lower', aspect='auto')
ax2.set_title("Smoothing over 16 neighbors")
ax3.imshow(heatmap_32, extent=extent_32, origin='lower', aspect='auto')
ax3.set_title("Smoothing over 32 neighbors")
#Make the heatmap using a smoothing over 64 neighbors
ax4.imshow(heatmap_64, extent=extent_64, origin='lower', aspect='auto')
ax4.set_title("Smoothing over 64 neighbors")
plt.show()
which produces the following image:
As you see, the images look pretty nice, and we are able to identify different substructures on it. These images are constructed spreading a given weight for every point within a certain domain, defined by the smoothing length, which in turns is given by the distance to the closer nb neighbor (I've chosen 16, 32 and 64 for the examples). So, higher density regions typically are spread over smaller regions compared to lower density regions.
The function myplot is just a very simple function that I've written in order to give the x,y data to py-sphviewer to do the magic.

If you are using 1.2.x
import numpy as np
import matplotlib.pyplot as plt
x = np.random.randn(100000)
y = np.random.randn(100000)
plt.hist2d(x,y,bins=100)
plt.show()

Seaborn now has the jointplot function which should work nicely here:
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
# Generate some test data
x = np.random.randn(8873)
y = np.random.randn(8873)
sns.jointplot(x=x, y=y, kind='hex')
plt.show()

Here's Jurgy's great nearest neighbour approach but implemented using scipy.cKDTree. In my tests it's about 100x faster.
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.cm as cm
from scipy.spatial import cKDTree
def data_coord2view_coord(p, resolution, pmin, pmax):
dp = pmax - pmin
dv = (p - pmin) / dp * resolution
return dv
n = 1000
xs = np.random.randn(n)
ys = np.random.randn(n)
resolution = 250
extent = [np.min(xs), np.max(xs), np.min(ys), np.max(ys)]
xv = data_coord2view_coord(xs, resolution, extent[0], extent[1])
yv = data_coord2view_coord(ys, resolution, extent[2], extent[3])
def kNN2DDens(xv, yv, resolution, neighbours, dim=2):
"""
"""
# Create the tree
tree = cKDTree(np.array([xv, yv]).T)
# Find the closest nnmax-1 neighbors (first entry is the point itself)
grid = np.mgrid[0:resolution, 0:resolution].T.reshape(resolution**2, dim)
dists = tree.query(grid, neighbours)
# Inverse of the sum of distances to each grid point.
inv_sum_dists = 1. / dists[0].sum(1)
# Reshape
im = inv_sum_dists.reshape(resolution, resolution)
return im
fig, axes = plt.subplots(2, 2, figsize=(15, 15))
for ax, neighbours in zip(axes.flatten(), [0, 16, 32, 63]):
if neighbours == 0:
ax.plot(xs, ys, 'k.', markersize=5)
ax.set_aspect('equal')
ax.set_title("Scatter Plot")
else:
im = kNN2DDens(xv, yv, resolution, neighbours)
ax.imshow(im, origin='lower', extent=extent, cmap=cm.Blues)
ax.set_title("Smoothing over %d neighbours" % neighbours)
ax.set_xlim(extent[0], extent[1])
ax.set_ylim(extent[2], extent[3])
plt.savefig('new.png', dpi=150, bbox_inches='tight')

and the initial question was... how to convert scatter values to grid values, right?
histogram2d does count the frequency per cell, however, if you have other data per cell than just the frequency, you'd need some additional work to do.
x = data_x # between -10 and 4, log-gamma of an svc
y = data_y # between -4 and 11, log-C of an svc
z = data_z #between 0 and 0.78, f1-values from a difficult dataset
So, I have a dataset with Z-results for X and Y coordinates. However, I was calculating few points outside the area of interest (large gaps), and heaps of points in a small area of interest.
Yes here it becomes more difficult but also more fun. Some libraries (sorry):
from matplotlib import pyplot as plt
from matplotlib import cm
import numpy as np
from scipy.interpolate import griddata
pyplot is my graphic engine today,
cm is a range of color maps with some initeresting choice.
numpy for the calculations,
and griddata for attaching values to a fixed grid.
The last one is important especially because the frequency of xy points is not equally distributed in my data. First, let's start with some boundaries fitting to my data and an arbitrary grid size. The original data has datapoints also outside those x and y boundaries.
#determine grid boundaries
gridsize = 500
x_min = -8
x_max = 2.5
y_min = -2
y_max = 7
So we have defined a grid with 500 pixels between the min and max values of x and y.
In my data, there are lots more than the 500 values available in the area of high interest; whereas in the low-interest-area, there are not even 200 values in the total grid; between the graphic boundaries of x_min and x_max there are even less.
So for getting a nice picture, the task is to get an average for the high interest values and to fill the gaps elsewhere.
I define my grid now. For each xx-yy pair, i want to have a color.
xx = np.linspace(x_min, x_max, gridsize) # array of x values
yy = np.linspace(y_min, y_max, gridsize) # array of y values
grid = np.array(np.meshgrid(xx, yy.T))
grid = grid.reshape(2, grid.shape[1]*grid.shape[2]).T
Why the strange shape? scipy.griddata wants a shape of (n, D).
Griddata calculates one value per point in the grid, by a predefined method.
I choose "nearest" - empty grid points will be filled with values from the nearest neighbor. This looks as if the areas with less information have bigger cells (even if it is not the case). One could choose to interpolate "linear", then areas with less information look less sharp. Matter of taste, really.
points = np.array([x, y]).T # because griddata wants it that way
z_grid2 = griddata(points, z, grid, method='nearest')
# you get a 1D vector as result. Reshape to picture format!
z_grid2 = z_grid2.reshape(xx.shape[0], yy.shape[0])
And hop, we hand over to matplotlib to display the plot
fig = plt.figure(1, figsize=(10, 10))
ax1 = fig.add_subplot(111)
ax1.imshow(z_grid2, extent=[x_min, x_max,y_min, y_max, ],
origin='lower', cmap=cm.magma)
ax1.set_title("SVC: empty spots filled by nearest neighbours")
ax1.set_xlabel('log gamma')
ax1.set_ylabel('log C')
plt.show()
Around the pointy part of the V-Shape, you see I did a lot of calculations during my search for the sweet spot, whereas the less interesting parts almost everywhere else have a lower resolution.

Make a 2-dimensional array that corresponds to the cells in your final image, called say heatmap_cells and instantiate it as all zeroes.
Choose two scaling factors that define the difference between each array element in real units, for each dimension, say x_scale and y_scale. Choose these such that all your datapoints will fall within the bounds of the heatmap array.
For each raw datapoint with x_value and y_value:
heatmap_cells[floor(x_value/x_scale),floor(y_value/y_scale)]+=1

Very similar to #Piti's answer, but using 1 call instead of 2 to generate the points:
import numpy as np
import matplotlib.pyplot as plt
pts = 1000000
mean = [0.0, 0.0]
cov = [[1.0,0.0],[0.0,1.0]]
x,y = np.random.multivariate_normal(mean, cov, pts).T
plt.hist2d(x, y, bins=50, cmap=plt.cm.jet)
plt.show()
Output:

Here's one I made on a 1 Million point set with 3 categories (colored Red, Green, and Blue). Here's a link to the repository if you'd like to try the function. Github Repo
histplot(
X,
Y,
labels,
bins=2000,
range=((-3,3),(-3,3)),
normalize_each_label=True,
colors = [
[1,0,0],
[0,1,0],
[0,0,1]],
gain=50)

I'm afraid I'm a little late to the party but I had a similar question a while ago. The accepted answer (by #ptomato) helped me out but I'd also want to post this in case it's of use to someone.
''' I wanted to create a heatmap resembling a football pitch which would show the different actions performed '''
import numpy as np
import matplotlib.pyplot as plt
import random
#fixing random state for reproducibility
np.random.seed(1234324)
fig = plt.figure(12)
ax1 = fig.add_subplot(121)
ax2 = fig.add_subplot(122)
#Ratio of the pitch with respect to UEFA standards
hmap= np.full((6, 10), 0)
#print(hmap)
xlist = np.random.uniform(low=0.0, high=100.0, size=(20))
ylist = np.random.uniform(low=0.0, high =100.0, size =(20))
#UEFA Pitch Standards are 105m x 68m
xlist = (xlist/100)*10.5
ylist = (ylist/100)*6.5
ax1.scatter(xlist,ylist)
#int of the co-ordinates to populate the array
xlist_int = xlist.astype (int)
ylist_int = ylist.astype (int)
#print(xlist_int, ylist_int)
for i, j in zip(xlist_int, ylist_int):
#this populates the array according to the x,y co-ordinate values it encounters
hmap[j][i]= hmap[j][i] + 1
#Reversing the rows is necessary
hmap = hmap[::-1]
#print(hmap)
im = ax2.imshow(hmap)
Here's the result

None of these solutions worked for my application, so this is what I came up with. Essentially I am placing a 2D Gaussian at every single point:
import cv2
import numpy as np
import matplotlib.pyplot as plt
def getGaussian2D(ksize, sigma, norm=True):
oneD = cv2.getGaussianKernel(ksize=ksize, sigma=sigma)
twoD = np.outer(oneD.T, oneD)
return twoD / np.sum(twoD) if norm else twoD
def pt2heat(pts, shape, kernel=16, sigma=5):
heat = np.zeros(shape)
k = getGaussian2D(kernel, sigma)
for y,x in pts:
x, y = int(x), int(y)
for i in range(-kernel//2, kernel//2):
for j in range(-kernel//2, kernel//2):
if 0 <= x+i < shape[0] and 0 <= y+j < shape[1]:
heat[x+i, y+j] = heat[x+i, y+j] + k[i+kernel//2, j+kernel//2]
return heat
heat = pts2heat(pts, img.shape[:2])
plt.imshow(heat, cmap='heat')
Here are the points overlayed ontop of it's associated image, along with the resulting heat map: