I am generating a groundwater elevation contour with Matplotlib. See below
Here is what I have now; how can I add water flow arrows like the image below?
I want to add arrows to make it look like this:
If anyone has some ideas and/or code samples that would be greatly appreciated.
You'll need a recent (>= 1.2) version of matplotlib, but streamplot does this. You just need to take the negative gradient of your head (a.k.a. "water table" for surface aquifers) grid.
As a quick example generated from random point observations of head:
import numpy as np
from scipy.interpolate import Rbf
import matplotlib.pyplot as plt
# Make data repeatable
np.random.seed(1981)
# Generate some random wells with random head (water table) observations
x, y, z = np.random.random((3, 10))
# Interpolate these onto a regular grid
xi, yi = np.mgrid[0:1:100j, 0:1:100j]
func = Rbf(x, y, z, function='linear')
zi = func(xi, yi)
# -- Plot --------------------------
fig, ax = plt.subplots()
# Plot flowlines
dy, dx = np.gradient(-zi.T) # Flow goes down gradient (thus -zi)
ax.streamplot(xi[:,0], yi[0,:], dx, dy, color='0.8', density=2)
# Contour gridded head observations
contours = ax.contour(xi, yi, zi, linewidths=2)
ax.clabel(contours)
# Plot well locations
ax.plot(x, y, 'ko')
plt.show()
Related
I'm trying to create a piecewise linear interpolation routine and I'm pretty new to all of this so I'm very uncertain of what needs to be done.
I've generate a set of data points in 3D which gives variation in all 3 directions. I want to interpolate between these data points and plot in 3D.
The current data set is much smaller than the final one will be. Linear interpolation is important.
here's the current code
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
import scipy.interpolate as interp
x = np.linspace(-1.3,1.3,10)
y1 = np.linspace(.5,0.,5)
y2 = np.linspace(0.,.5,5)
y = np.hstack((y1,y2))
z1 = np.linspace(.1,0.,5)
z2 = np.linspace(0.,.1,5)
z = np.hstack((z1,z2))
data = np.dstack([x,y,z])
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
f = interp.interp2d(x, y, z, kind='linear')
xnew = np.linspace(-1.3,1.3,100)
y1new = np.linspace(.5,0.,50)
y2new = np.linspace(0.,.5,50)
ynew = np.hstack((y1new,y2new))
znew = f(xnew,ynew)
ax.plot(x,y,znew, 'b-')
ax.scatter(x,y,z,'ro')
plt.show()
As I said, dataset is just to add variation. The real set will be much bigger but have less variation. I don't really understand the interpolation tool and the scipy documentation isn't very clear
would appreciate suggestions
2D ok. Please help with 3D
What I'm trying to do is build something that takes data points for deflections of a beam an interpolates between the data points. I wanted to to this in 3D and get a 3D plot showing the deflection along the x-axis in both y and z directions at the same time. As a stop gap measure I've used the below code to individually show deflection in y dir and z dir. Note, the data set is randomly generated for the moment. Some choices might look strange at the mo, but that's to sorta stick to the kinda range the final data set will use. The code below works for a 2D system so may be helpful to someone. I'd still really appreciate if someone could help me do this in 3D.
import numpy as np
import matplotlib.pyplot as plt
from scipy.interpolate import CubicSpline
u=10
x = np.linspace(-1.3,1.3,u) #regular x-data
y = np.random.random_sample(u)/4 #random y data
z = np.random.random_sample(u)/10 # random zdata
ynone = np.ones(u)*0.1 #no deflection dataset
znone = np.ones(u)*0.05
xspace = np.linspace(-1.3, 1.3, u*100)
ydefl = CubicSpline(x, y) #creating cubinc spline function for original data
zdefl = CubicSpline(x, z)
plt.subplot(2, 1, 1)
plt.plot(x, ynone, '-',label='y - no deflection')
plt.plot(x, y, 'go',label='y-deflection data')
plt.plot(xspace, ydefl(xspace), label='spline') #plot xspace vs spline function of xspace
plt.title('X [m]s')
plt.ylabel('Y [m]')
plt.legend(loc='best', ncol=3)
plt.subplot(2, 1, 2)
plt.plot(x, znone, '-',label='z - no deflection')
plt.plot(x, z, 'go',label='z-deflection data')
plt.plot(xspace, zdefl(xspace),label='spline')
plt.xlabel('X [m]')
plt.ylabel('Z [m]')
plt.legend(loc='best', ncol=3)
plt.show()
I got a problem when I was plotting a 3d figure using matplotlib of python. Using the following python function, I got this figure:
Here X, Y are meshed grids and Z and Z_ are functions of X and Y. C stands for surface color.
import numpy as np
from mpl_toolkits.mplot3d import Axes3D
from matplotlib import cm
import matplotlib.pyplot as plt
def plot(X, Y, Z, Z_, C):
fig = plt.figure()
ax = fig.gca(projection='3d')
surf = ax.plot_surface(
X, Y, Z, rstride=1, cstride=1,
facecolors=cm.jet(C),
linewidth=0, antialiased=False, shade=False)
surf_ = ax.plot_surface(
X, Y, Z_, rstride=1, cstride=1,
facecolors=cm.jet(C),
linewidth=0, antialiased=False, shade=False)
ax.view_init(elev=7,azim=45)
plt.show()
But now I want to cut this figure horizontally and only the part whose z is between -1 and 2 remain.
What I want, plotted with gnuplot, is this:
I have tried ax.set_zlim3d and ax.set_zlim, but neither of them give me the desired figure. Does anybody know how to do it using python?
Nice conical intersections you have there:)
What you're trying to do should be achieved by setting the Z data you want to ignore to NaN. Using graphene's tight binding band structure as an example:
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
# generate dummy data (graphene tight binding band structure)
kvec = np.linspace(-np.pi,np.pi,101)
kx,ky = np.meshgrid(kvec,kvec)
E = np.sqrt(1+4*np.cos(3*kx/2)*np.cos(np.sqrt(3)/2*ky) + 4*np.cos(np.sqrt(3)/2*ky)**2)
# plot full dataset
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
ax.plot_surface(kx,ky,E,cmap='viridis',vmin=-E.max(),vmax=E.max(),rstride=1,cstride=1)
ax.plot_surface(kx,ky,-E,cmap='viridis',vmin=-E.max(),vmax=E.max(),rstride=1,cstride=1)
# focus on Dirac cones
Elim = 1 #threshold
E[E>Elim] = np.nan
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
#ax.plot_surface(kx2,ky2,E2,cmap='viridis',vmin=-Elim,vmax=Elim)
#ax.plot_surface(kx2,ky2,-E2,cmap='viridis',vmin=-Elim,vmax=Elim)
ax.plot_surface(kx,ky,E,cmap='viridis',rstride=1,cstride=1,vmin=-Elim,vmax=Elim)
ax.plot_surface(kx,ky,-E,cmap='viridis',rstride=1,cstride=1,vmin=-Elim,vmax=Elim)
plt.show()
The results look like this:
Unfortunately, there are problems with the rendering of the second case: the apparent depth order of the data is messed up in the latter case: cones in the background are rendered in front of the front ones (this is much clearer in an interactive plot). The problem is that there are more holes than actual data, and the data is not connected, which confuses the renderer of plot_surface. Matplotlib has a 2d renderer, so 3d visualization is a bit of a hack. This means that for complex overlapping surfaces you'll more often than not get rendering artifacts (in particular, two simply connected surfaces are either fully behind or fully in front of one another).
We can get around the rendering bug by doing a bit more work: keeping the data in a single surface by not using nans, but instead colouring the the surface to be invisible where it doesn't interest us. Since the surface we're plotting now includes the entire original surface, we have to set the zlim manually in order to focus on our region of interest. For the above example:
from matplotlib.cm import get_cmap
# create a color mapping manually
Elim = 1 #threshold
cmap = get_cmap('viridis')
colors_top = cmap((E + Elim)/2/Elim) # listed colormap that maps E from [-Elim, Elim] to [0.0, 1.0] for color mapping
colors_bott = cmap((-E + Elim)/2/Elim) # same for -E branch
colors_top[E > Elim, -1] = 0 # set outlying faces to be invisible (100% transparent)
colors_bott[-E < -Elim, -1] = 0
# in nature you would instead have something like this:
#zmin,zmax = -1,1 # where to cut the _single_ input surface (x,y,z)
#cmap = get_cmap('viridis')
#colors = cmap((z - zmin)/(zmax - zmin))
#colors[(z < zmin) | (z > zmax), -1] = 0
# then plot_surface(x, y, z, facecolors=colors, ...)
# or for your specific case where you have X, Y, Z and C:
#colors = get_cmap('viridis')(C)
#colors[(z < zmin) | (z > zmax), -1] = 0
# then plot_surface(x, y, z, facecolors=colors, ...)
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
# pass the mapped colours as the facecolors keyword arg
s1 = ax.plot_surface(kx, ky, E, facecolors=colors_top, rstride=1, cstride=1)
s2 = ax.plot_surface(kx, ky, -E, facecolors=colors_bott, rstride=1, cstride=1)
# but now we need to manually hide the invisible part of the surface:
ax.set_zlim(-Elim, Elim)
plt.show()
Here's the output:
Note that it looks a bit different from the earlier figures because 3 years have passed in between and the current version of matplotlib (3.0.2) has very different (and much prettier) default styles. In particular, edges are now transparent in surface plots. But the main point is that the rendering bug is gone, which is evident if you start rotating the surface around in an interactive plot.
I am trying to make a contour plot in python with complex numbers (i am using matplotlib, pylab).
I am working with sharp bounds on harmonic polynomials, but specifically right now I am trying to plot:
Re(z(bar) - e^(z))= 0
Im(z(bar) - e^z) = 0
and plot them over each other in a contour in order to find their zeros to determine how many solutions there are to the equation z(bar) = e^(z).
Does anyone have experience in contour plotting, specifically with complex numbers?
import numpy as np
from matplotlib import pyplot as plt
x = np.r_[0:10:30j]
y = np.r_[0:10:20j]
X, Y = np.meshgrid(x, y)
Z = X*np.exp(1j*Y) # some arbitrary complex data
def plotit(z, title):
plt.figure()
cs = plt.contour(X,Y,z) # contour() accepts complex values
plt.clabel(cs, inline=1, fontsize=10) # add labels to contours
plt.title(title)
plt.savefig(title+'.png')
plotit(Z, 'real')
plotit(Z.real, 'explicit real')
plotit(Z.imag, 'imaginary')
plt.show()
EDIT: Above is my code, and note that for Z, I need to plot both real and imaginary parts of (x- iy) - e^(x+iy)=0. The current Z that is there is simply arbitrary. It is giving me an error for not having a 2D array when I try to plug mine in.
I don't know how you are plotting since you didn't post any code, but in general I advise moving away from using pylab or the pyplot interface to matplotlib, using the direct object methods is much more robust and just as simple. Here is an example of plotting contours of two sets of data on the same plot.
import numpy as np
import matplotlib.pyplot as plt
# making fake data
x = np.linspace(0, 2)
y = np.linspace(0, 2)
c = x[:,np.newaxis] * y
c2 = np.flipud(c)
# plot
fig, ax = plt.subplots(1, 1)
cont1 = ax.contour(x, y, c, colors='b')
cont2 = ax.contour(x, y, c2, colors='r')
cont1.clabel()
cont2.clabel()
plt.show()
For tom10, here is the plot this code produces. Note that setting colors to a single color makes distinguishing the two plots much easier.
I'm trying to plot some data for a measurement taken from between two surfaces. The z-direction in the system is defined as normal to the surfaces. The problem is that along the x-axis of my plot I'm varying the separation distance between the two surfaces which means that for every slice, the min/max of the y-axis change. I've sort circumvented this by presenting a normalized y-axis where z_min is the bottom surface and z_max is the top surface:
However, this representation somewhat distorts the data. Ideally I would like to show the actual distance to the wall on the y-axis and just leave the areas outside of the system bounds white. I (poorly) sketched what I'm envisioning here (the actual distribution on the heatmap should look different, of course):
I can pretty easily plot what I want as a 3D scatter plot like so:
But how do I get the data into a plot-able form for a heatmap?
I'm guessing I would have to blow up the MxN array and fill in missing values through interpolation or simply mark them as NAN? But then I'm also not quite sure how to add a hard cutoff to my color scheme to make everything outside of the system white.
You can do this with pcolormesh which takes the corners of quadrilaterals as the arguements
X, Y = np.meshgrid(np.linspace(0, 10, 100), np.linspace(0, 2*np.pi, 150),)
h = np.sin(Y)
Y *= np.linspace(.5, 1, 100)
fig, ax = plt.subplots(1, 1)
ax.pcolormesh(X, Y, h)
Below an implementation with triangular mesh contouring, based on CT Zhu example.
If your domain is not convex, you will need to provide your own triangles to the triangulation, as default Delaunay triangulation meshes the convex hull from your points.
import matplotlib
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.tri as mtri
y = np.array([np.linspace(-i, i, 51) for i in (
np.linspace(5, 10))[::-1]])
x = (np.zeros((50, 51)) +
np.linspace(1, 6, 50)[..., np.newaxis])
z = (np.zeros((50, 51)) -
np.linspace(-5, 5, 51)**2 + 10) # make up some z data
x = x.flatten()
y = y.flatten()
z = z.flatten()
print "x shape: ", x.shape
triang = mtri.Triangulation(x, y)
plt.tricontourf(triang, z)
plt.colorbar()
plt.show()
I guess, maybe 2d interpolation by using griddata will be what you want?
from matplotlib.mlab import griddata
xi=linspace(1,5,100)
yi=linspace(-10.5, 10.5, 100)
y=array([linspace(-i, i, 51) for i in (linspace(5,10))[::-1]]) #make up some y vectors with different range
x=zeros((50,51))+linspace(1,6, 50)[...,newaxis]
z=zeros((50,51))-linspace(-5, 5,51)**2+10 #make up some z data
x=x.flatten()
y=y.flatten()
z=z.flatten()
zi=griddata(x, y, z, xi, yi)
plt.contourf(xi, yi, zi, levels=-linspace(-5, 5,51)**2+10)
I am trying to model an asteroid using plot_surface and plot_wireframe. I have x y and z values for the points on the surface of the asteroid. The wireframe is accurate to the shape of the asteroid but the surface plot does not fit the wireframe. How can I get the surface plot to fit the wireframe or how could i use the wireframe to get a 3d solid model? Here is my code for the model:
from mpl_toolkits.mplot3d import axes3d
import matplotlib.pyplot as plt
import numpy as np
from matplotlib import cm
from matplotlib.mlab import griddata
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
data = np.genfromtxt('data.txt')
x = data[:,0]
y = data[:,1]
z = data[:,2]
ax.plot_wireframe(x, y, z, rstride=1, cstride=1, alpha=1)
xi = np.linspace(min(x), max(x))
yi = np.linspace(min(y), max(y))
X, Y = np.meshgrid(xi, yi)
Z = griddata(x, y, z, xi, yi)
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)
plt.show()
The data is in this format, although there are a lot more lines in the original file:
-1.7738946051191869E-002 4.3461451610545973E-002 1.3393057231408241
-0.29733561550902488 0.32305812106837900 1.3393057231408241
-0.29733561550902488 0.16510132228266330 1.3548631099230350
-0.21872587865015569 2.4170900455101410E-002 1.3610011616437809
1.4452975249810950E-002 -0.20900795344486520 1.3610011616437809
1.5732454381265970E-002 -0.20900795344486520 1.3608751439485580
-0.34501536374240321 0.51320241386595655 1.3158820995876130
-0.40193014435941982 0.45628763324893978 1.3158820995876130
-0.42505849480150409 0.28183419537116011 1.3307863198123011
-0.18994178462386799 -0.19294290416565860 1.3424523041534830
1.4452975249810939E-002 -0.39733766403933751 1.3424523041534830
5.8021940902131752E-002 -0.57108837516584876 1.3210481842104100
9.3746267961881152E-002 -0.61017602710257668 1.3136798474111200
0.26609469681891229 -0.43782759824554562 1.3136798474111200
0.17938460413447810 0.39179924148155021 1.2357401964919650
8.9613011902522258E-002 0.42818009222325598 1.2584008460875080
0.33671539027096409 -0.47165177581327772 1.2965073126705291
0.53703772594296528 -0.47165177581327777 1.2357401964919561
-0.19242375014122229 0.71021685426700043 1.2584008460875080
-0.34501536374240321 0.66763766324752027 1.2904902860951690
Hope you can help
Could you provide an image of what is happening? I'm guessing that you are getting an image where the surface of the asteroid seems to be jumping all over the place. Is that correct? If so it could be caused by the plotter not knowing the order of the points.
If we have a set of points, which contains all the points of a unit circle, then we would expect a drawing of those points to create a unit circle. If, however, you decided to connect each point to two other points, then it wouldn't necessarily look like a circle. If (for some reason) you connected one point to another point on the other side of the circle and continued to do that until each point was connected to two other points, it may or may not look like a circle because each point isn't necessarily connected to the adjacent points.
The same is true for the your asteroid. You need to come up with some scheme so that the plotter knows how to connect the points, otherwise you will continue to have the same problem.
The following example of a circle should illustrate my point:
import math
import matplotlib.pylab as plt
import random
thetaList = range(360)
random.shuffle(thetaList)
degToRad = lambda x: float(x) * math.pi / float(180)
x = [math.cos(degToRad(theta)) for theta in thetaList]
y = [math.sin(degToRad(theta)) for theta in thetaList]
#plot the cirlce
plt.plot(x,y)
plt.show()