def viz_dimer_positions_ipv(positions,
size=5,
cmap_name="tab20c",
color_feature_name=None):
try:
import ipyvolume as ipv
except ImportError:
mess = "YOu need to install the ipyvolume library. "
mess += "Please follow the official instructions at https://github.com/maartenbreddels/ipyvolume."
raise ImportError(mess)
# Only show visible dimers
selected_dimers = positions[positions["visible"]]
x, y, z = selected_dimers[["x", "y", "z"]].values.astype("float").T
if color_feature_name:
# TODO: that code should be much simpler...
cmap = matplotlib.cm.get_cmap(cmap_name)
categories = selected_dimers[color_feature_name].unique()
color_indices = cmap([i / len(categories) for i in categories])
color = np.zeros((len(selected_dimers[color_feature_name]), 4))
for color_index, _ in enumerate(categories):
mask = selected_dimers[color_feature_name] == categories[
color_index]
color[mask] = color_indices[color_index]
else:
color = "#e4191b"
ipv.figure(height=800, width=1000)
ipv.scatter(x, y, z, size=size, marker="sphere", color=color)
ipv.squarelim()
ipv.show()
def plot_clusters(self, threshold, rcut=28, size=5.):
ipv.clear()
s = self.trace <= threshold
xyz = self.stress_coord[s]
x, y, z = xyz[:, 0], xyz[:, 1], xyz[:, 2]
clustering = DBSCAN(eps=rcut, min_samples=1).fit(xyz)
labels = np.unique(clustering.labels_)
counts, edges = np.histogram(clustering.labels_,
bins=np.arange(labels.min(),
labels.max() + 1))
largest = edges[counts.argmax()]
color_dict = {}
for l in labels:
color_dict[l] = tuple(np.random.uniform(0, 1, size=3))
color_dict[largest] = (1, 1, 1)
colors = np.array([color_dict[c] for c in clustering.labels_])
color = np.array([(x - x.min()) / x.ptp(),
np.ones_like(x),
np.ones_like(x)]).T
ipv.scatter(x, y, z, size=size, marker="sphere", color=colors)
ipv.show()
def complete(final_solution: Solution,
correct_solution: Solution = None) -> None:
"""
Will print result in 3D world
:return: Noting
"""
# init of plot-output #1
figure = plot.figure()
axes = figure.add_subplot(111, projection='3d')
for part in final_solution.partitions:
color = random.choice(list_of_colors)
marker = random.choice(list_of_marker)
for p in part:
axes.scatter(p.x, p.y, p.z, marker=marker, c=color)
axes.set_xlabel("X Achse")
axes.set_ylabel("Y Achse")
axes.set_zlabel("Z Achse")
plot.show()
# init 3D view
ipv.current.containers.clear()
ipv.current.figures.clear()
extra_figures = list()
# create correct solution
if correct_solution is not None:
figure_correct = ipv.figure("correct")
ipv.pylab.xyzlim(-1, 11)
for part in correct_solution.partitions:
marker = random.choice(list_of_IPYVmarker)
color = random.choice(list_of_colors)
ipv.scatter(*convert.partition_to_IpyVolume(part),
marker=marker,
color=color)
extra_figures.append(figure_correct)
figure_result = ipv.figure("result")
container = ipv.gcc()
ipv.current.container = widgets.HBox(container.children)
ipv.current.containers["result"] = ipv.current.container
# crate computed solution
for part in final_solution.partitions:
marker = random.choice(list_of_IPYVmarker)
color = random.choice(list_of_colors)
ipv.scatter(*convert.partition_to_IpyVolume(part),
marker=marker,
color=color)
ipv.pylab.xyzlim(-1, 11)
ipv.current.container.children = list(
ipv.current.container.children) + extra_figures
ipv.show()
# save in a separate file
ipv.save("example.html")
# destroy 3D views
ipv.clear()
def FCC(L_x, L_y, L_z, Color='red'):
Ax = []
Ay = []
Az = []
#for each layer of the atom in the x direction. the *2-1 is so that L_x is the number of unit cells
for O in range(L_x * 2 - 1):
#for each layer in the y direction.
for S in range(L_y * 2 - 1):
#for each layer in the z direction.
for A in range(L_z * 2 - 1):
if (O + S + A) % 2 == 0:
#the 0.5 is becasue L_xyz was multiplied by 2
Ax.append(float(O * .5))
Ay.append(float(S * .5))
Az.append(float(A * .5))
elif S % 2 == 1 and O % 2 == 1 and A % 2 == 0:
Ax.append(float(O * .5))
Ay.append(float(S * .5))
Az.append(float(A * .5))
#if there should not be an atom
#in theory this could be used to remove other atoms to make cubic or BCC structure
#the x, y, z limits
ipv.xyzlim(0, 10) #define as a function highest of L_xyz
ipv.scatter(np.array(Ax),
np.array(Ay),
np.array(Az),
marker='sphere',
size=6.5,
color=Color)
ipv.show()
def show_colorspace(cspace: np.array,
clip=True,
size=0.5,
marker='sphere',
**kwargs) -> None:
"""
Visualise colorspace vector as an interactive 3D figure. Colorspace can have extra columns
By default RGB channels are clipped to the range [0,1].
Extra arguments can be used to control the appearance of ipyvolume.scatter
"""
assert isinstance(cspace, DataFrame), "Colorspace must be a dataframe"
assert all(np.isin(['R', 'G', 'B'],
cspace.columns)), "Colorspace must contain RGB columns"
fig = ipv.figure()
if clip:
ipv.scatter(cspace.loc[:, 'R'].values,
cspace.loc[:, 'G'].values,
cspace.loc[:, 'B'].values,
color=np.clip(cspace[['R', 'G', 'B']].values, 0, 1),
s=size,
marker=marker,
*kwargs)
else:
ipv.scatter(cspace.loc[:, 'R'].values,
cspace.loc[:, 'G'].values,
cspace.loc[:, 'B'].values,
color=cspace[['R', 'G', 'B']].values,
s=size,
marker=marker,
*kwargs)
ipv.show()
def viz_dimer_positions(positions,
size=5,
cmap_name="tab20c",
color_feature_name=None):
import ipyvolume as ipv
# Only show visible dimers
selected_dimers = positions[positions['visible'] is True]
x, y, z = selected_dimers[['x', 'y', 'z']].values.astype('float').T
if color_feature_name:
# TODO: that code should be much simpler...
cmap = matplotlib.cm.get_cmap(cmap_name)
categories = selected_dimers[color_feature_name].unique()
color_indices = cmap([i / len(categories) for i in categories])
color = np.zeros((len(selected_dimers[color_feature_name]), 4))
for color_index in enumerate(categories):
color[selected_dimers[color_feature_name] ==
categories[color_index]] = color_indices[color_index]
else:
color = '#e4191b'
ipv.figure(height=800, width=1000)
ipv.scatter(x, y, z, size=size, marker='sphere', color=color)
ipv.squarelim()
ipv.show()
def PlotSkeleton(seg_name, plot_type='node', out_res=(30, 48, 48)):
print('Read skeletons')
skeletons = ReadSkeletons(seg_name,
skeleton_algorithm='thinning',
downsample_resolution=out_res,
read_edges=True)
print('Plot skeletons')
node_list = skeletons[1].get_nodes()
nodes = np.stack(node_list).astype(float)
junction_idx = skeletons[1].get_junctions()
junctions = nodes[junction_idx, :]
ends = skeletons[1].get_ends()
jns_ends = np.vstack([junctions, ends])
IX, IY, IZ = 2, 1, 0
ipv.figure()
if plot_type == 'nodes':
nodes = ipv.scatter(nodes[:,IX], nodes[:,IY], nodes[:,IZ], \
size=0.5, marker='sphere', color='blue')
elif plot_type == 'edges':
edges = skeletons[1].get_edges().astype(float)
for e1, e2 in edges:
if not ((e1[IX] == e2[IX]) and (e1[IY] == e2[IY]) and
(e1[IZ] == e2[IZ])):
ipv.plot([e1[IX], e2[IX]], [e1[IY], e2[IY]], [e1[IZ], e2[IZ]], \
color='blue')
jns = ipv.scatter(jns_ends[:,IX], jns_ends[:,IY], jns_ends[:,IZ], \
size=0.85, marker='sphere', color='red')
ipv.pylab.style.axes_off()
ipv.pylab.style.box_off()
ipv.save(seg_name + '_skel.html')
def _display_sphere(
self,
fluxes,
distances,
theta,
phi,
cmap="magma",
distance_transform=None,
use_log=False,
fig=None,
background_color="white",
show=True,
**kwargs,
):
if len(fluxes) == 0:
log.warning("There are no detections to display")
return
if fig is None:
fig = ipv.figure()
ipv.pylab.style.box_off()
ipv.pylab.style.axes_off()
ipv.pylab.style.set_style_dark()
ipv.pylab.style.background_color(background_color)
if distance_transform is not None:
distance = distance_transform(distances)
else:
distance = distances
x, y, z = _create_sphere_variables(self._r_max, distance, theta, phi)
_, colors = array_to_cmap(fluxes, cmap, use_log=True)
ipv.scatter(x, y, z, color=colors, marker="sphere", **kwargs)
r_value = fig
if show:
ipv.xyzlim(self._r_max)
fig.camera.up = [1, 0, 0]
control = pythreejs.OrbitControls(controlling=fig.camera)
fig.controls = control
control.autoRotate = True
fig.render_continuous = True
# control.autoRotate = True
# toggle_rotate = widgets.ToggleButton(description="Rotate")
# widgets.jslink((control, "autoRotate"), (toggle_rotate, "value"))
# r_value = toggle_rotate
return fig
def plot_geometry(self,
plot_e_r=True,
plot_e_phi=True,
plot_e_theta=True,
plot_bezier=True,
bezier_order=3):
"""Generates 3D ipyvolume plot
"""
fig = ipv.figure()
scatter = ipv.scatter(self.center[:, 0],
self.center[:, 1],
self.center[:, 2],
color='#ff0000',
size=5)
if plot_e_r:
self._my_ipv_vectorplot(np.repeat(self.center,
len(self.points),
axis=0),
self.e_r,
length=1,
N=1000,
size=0.2,
color='#00ff00')
if plot_e_theta:
self._my_ipv_vectorplot(self.points,
self.e_theta,
length=0.05,
N=100,
size=0.2,
color='#ff0000')
if plot_e_phi:
self._my_ipv_vectorplot(self.points,
self.e_phi,
length=0.05,
N=100,
size=0.2,
color='#ff0000')
if plot_bezier:
assert 1 <= bezier_order <= 3
p = self._arrange_bezier_points(self.bezier_points,
order=bezier_order)
self._plot_bezier_curves(p, N=100, size=0.5, color='#ff00ff')
scatter = ipv.scatter(self.points[:, 0],
self.points[:, 1],
self.points[:, 2],
color='#0000bb',
size=1)
ipv.xlim(0, 1)
ipv.ylim(0, 1)
ipv.zlim(0, 1)
return fig
def plot_voxel(voxels_position, marker="box", color="green", size=2.0):
if len(voxels_position) > 0:
x, y, z = (voxels_position[:,
0], voxels_position[:,
1], voxels_position[:,
2])
ipyvolume.scatter(x, y, z, size=size, marker=marker, color=color)
def _plot_bezier_curves(list_p, N=10, **kwargs):
"""Plot neurites as bezier curves.
"""
curves = []
for t in np.linspace(0, 1, N):
tmp = volutils.bezier(t, list_p)
curves.append(tmp)
curves = np.concatenate(curves, axis=0)
ipv.scatter(curves[:, 0], curves[:, 1], curves[:, 2], **kwargs)
def ipv_draw_point_cloud(ipv, pts, colors, pt_size=10):
pts = pts.reshape((-1, 3))
colors = colors.reshape((-1, 3))
assert colors.shape[0] == pts.shape[0]
ipv.scatter(x=pts[:, 0],
y=pts[:, 1],
z=pts[:, 2],
color=colors.reshape(-1, 3),
marker='sphere',
size=pt_size)
def show_centerline(pred):
pred[pred == 255] = 0
mesh = np.meshgrid(*[range(pred.shape[i]) for i in range(1, 4)],
indexing='ij')
cents = mesh + pred[1:]
centdots = cents.transpose([1, 2, 3, 0])[pred[0] > 0]
ipv.figure()
ipv.scatter(centdots[:, 0], centdots[:, 1], centdots[:, 2], size=0.1)
# ipv.volshow(pred[0])
ipv.show()
def NACL(L_x, L_y=None, L_z=None, ColorNa='red', ColorCl='green'):
if L_y == None:
L_y = L_x
if L_z == None:
L_z = L_x
if L_x > L_y:
Max = L_x
else:
Max = L_y
if L_z > Max:
Max = L_z #defined as a function highest of L_xyz
#define a list of positions to append to
Clx = []
Cly = []
Clz = []
Nax = []
Nay = []
Naz = []
#for each layer of the atom in the x direction. the *2-1 is so that L_x is the number of unit cells
for O in range(L_x * 2 - 1):
#for each layer in the y direction.
for S in range(L_y * 2 - 1):
#for each layer in the z direction.
for A in range(L_z * 2 - 1):
#if in this position there shoud be a Cl atom
if (O + S + A) % 2 == 0:
#the 0.5 is becasue L_xyz was multiplied by 2
Clx.append(float(O * .5))
Cly.append(float(S * .5))
Clz.append(float(A * .5))
#if there should be a Na atom
elif (O + S + A) % 2 == 1:
Nax.append(float(O * .5))
Nay.append(float(S * .5))
Naz.append(float(A * .5))
#if there should not be an atom
#in theory this could be used to remove other atoms to make cubic or BCC structure
#the x, y, z limits
ipv.xyzlim(0, Max)
#the color and size the Na and Cl atoms
ipv.scatter(np.array(Clx),
np.array(Cly),
np.array(Clz),
marker='sphere',
size=6.5,
color=ColorCl)
ipv.scatter(
np.array(Nax),
np.array(Nay),
np.array(Naz),
marker='sphere',
size=4,
color=ColorNa,
)
ipv.show()
def ipv_plot_coils ( coil1, coil2, maxSamplesToPlot = 10000):
assert( type(coil1) == np.ndarray and type(coil2) == np.ndarray)
maxSamplesToPlot = min( ( coil1.shape[0], maxSamplesToPlot ) )
stride = np.max((1, coil1.shape[0]//maxSamplesToPlot))
print( '\t plotting data - stride = {} '.format( stride ) )
ipv.figure()
ipv.scatter( coil1[::stride,0], coil1[::stride,1], coil1[::stride,2], size = .5, marker = 'sphere', color = 'lightgreen')
ipv.scatter( coil2[::stride,0], coil2[::stride,1], coil2[::stride,2], size = .5, marker = 'sphere', color = 'purple')
ipv.pylab.squarelim()
ipv.show()
def ipv_plot_data(data, colorStack='purple', holdOnFlag=False, markerSize=.25):
if not holdOnFlag:
ipv.figure(
width=600, height=600
) # create a new figure by default, otherwise append to existing
ipv.scatter(data[:, 0],
data[:, 1],
data[:, 2],
size=markerSize,
marker='sphere',
color=colorStack)
if not holdOnFlag: ipv.show()
def initalize_hpo(nTimesteps,
nParticles,
nWorkers,
paramRanges,
nParameters=3,
plotFlag=True):
accuracies = np.zeros((nTimesteps, nParticles))
bestParticleIndex = {}
globalBestParticleParams = np.zeros((nTimesteps, 1, nParameters))
particles = np.zeros((nTimesteps, nParticles, nParameters))
velocities = np.zeros((nTimesteps, nParticles, nParameters))
particleBoostingRounds = np.zeros((nTimesteps, nParticles))
# initial velocity is one
velocities[0, :, :] = np.ones((nParticles, nParameters)) * .25
# randomly initialize particle colors
particleColors = np.random.uniform(size=(1, nParticles, 3))
# best initialized to middle [ fictional particle ] -- is this necessary
bestParticleIndex[0] = -1
# grid initialize particles
x = np.linspace(paramRanges[0][1], paramRanges[0][2], nWorkers)
y = np.linspace(paramRanges[1][1], paramRanges[1][2], nWorkers)
z = np.linspace(paramRanges[2][1], paramRanges[2][2], nWorkers)
xx, yy, zz = np.meshgrid(x, y, z, indexing='xy')
xS = xx.reshape(1, -1)[0]
yS = yy.reshape(1, -1)[0]
zS = zz.reshape(1, -1)[0]
# clip middle particles
particles[0, :, 0] = np.hstack([xS[-nWorkers**2:], xS[:nWorkers**2]])
particles[0, :, 1] = np.hstack([yS[-nWorkers**2:], yS[:nWorkers**2]])
particles[0, :, 2] = np.hstack([zS[-nWorkers**2:], zS[:nWorkers**2]])
if plotFlag:
ipv.figure()
ipv.scatter(particles[0, :, 0],
particles[0, :, 1],
particles[0, :, 2],
marker='sphere',
color=particleColors)
ipv.xlim(paramRanges[0][1], paramRanges[0][2])
ipv.ylim(paramRanges[1][1], paramRanges[1][2])
ipv.zlim(paramRanges[2][1], paramRanges[2][2])
ipv.show()
return particles, velocities, accuracies, bestParticleIndex, globalBestParticleParams, particleBoostingRounds, particleColors
def scatter_nd(xs: t.Tensor):
xs = full_detach(xs)
if xs.shape[1] == 1:
plt.hist(xs[:, 0], bins=100)
elif xs.shape[1] == 2:
plt.scatter(*xs.T, s=.1)
elif xs.shape[1] == 3:
ipv.clear()
ipv.scatter(*xs.T)
ipv.show()
else:
assert False, f"Wrong dimensionality: {xs.shape=}"
def ShowAnimation3D(self, size=15):
ipv.figure()
ipv.style.use("dark")
x_Part = []
y_Part = []
z_Part = []
for Part in self.Parts:
temp_x = Part.x
temp_y = Part.y
temp_z = Part.z
x_Part.append(temp_x)
y_Part.append(temp_y)
z_Part.append(temp_z)
x = combine(self.speed * 5, *x_Part)
y = combine(self.speed * 5, *y_Part)
z = combine(self.speed * 5, *z_Part)
u = ipv.scatter(x, y, z, marker="sphere", size=10, color="green")
ipv.animation_control(u, interval=100)
ipv.xyzlim(-size, size)
ipv.show()
def _my_ipv_vectorplot(xyz, uvw, N=10, length=1, **kwargs):
"""Generates a scatter plot of small points along a vector.
Args:
xyz: start point
uvw: direction
N: number of points rendered
length: length of the line
**kwargs: additional arguments of ipv.scatter
"""
points = []
for x in np.linspace(0, length, N):
tmp = xyz + uvw * x
points.append(tmp)
points = np.concatenate(points, axis=0)
ipv.scatter(points[:, 0], points[:, 1], points[:, 2], **kwargs)
def spherical_galaxy_orbit(orbit_x,
orbit_y,
orbit_z,
N_stars=100,
sigma_r=1,
orbit_visible=False,
orbit_line_interpolate=5,
N_star_orbits=10,
color=[255, 220, 200],
size_star=1,
scatter_kwargs={}):
"""Create a fake galaxy around the points orbit_x/y/z with N_stars around it"""
if orbit_line_interpolate > 1:
x = np.linspace(0, 1, len(orbit_x))
x_smooth = np.linspace(0, 1, len(orbit_x) * orbit_line_interpolate)
kind = 'quadratic'
orbit_x_line = scipy.interpolate.interp1d(x, orbit_x, kind)(x_smooth)
orbit_y_line = scipy.interpolate.interp1d(x, orbit_y, kind)(x_smooth)
orbit_z_line = scipy.interpolate.interp1d(x, orbit_z, kind)(x_smooth)
else:
orbit_x_line = orbit_x
orbit_y_line = orbit_y
orbit_z_line = orbit_z
line = ipv.plot(orbit_x_line,
orbit_y_line,
orbit_z_line,
visible=orbit_visible)
x = np.repeat(orbit_x, N_stars).reshape((-1, N_stars))
y = np.repeat(orbit_y, N_stars).reshape((-1, N_stars))
z = np.repeat(orbit_z, N_stars).reshape((-1, N_stars))
xr, yr, zr = np.random.normal(0, scale=sigma_r, size=(3, N_stars)) # +
r = np.sqrt(xr**2 + yr**2 + zr**2)
for i in range(N_stars):
a = np.linspace(0, 1, x.shape[0]) * 2 * np.pi * N_star_orbits
xo = r[i] * np.sin(a)
yo = r[i] * np.cos(a)
zo = a * 0
xo, yo, zo = np.dot(_randomSO3(), [xo, yo, zo])
#print(x.shape, xo.shape)
x[:, i] += xo
y[:, i] += yo
z[:, i] += zo
sprite = ipv.scatter(x,
y,
z,
texture=radial_sprite((64, 64), color),
marker='square_2d',
size=size_star,
**scatter_kwargs)
with sprite.material.hold_sync():
sprite.material.blending = pythreejs.BlendingMode.CustomBlending
sprite.material.blendSrc = pythreejs.BlendFactors.SrcColorFactor
sprite.material.blendDst = pythreejs.BlendFactors.OneFactor
sprite.material.blendEquation = 'AddEquation'
sprite.material.transparent = True
sprite.material.depthWrite = False
sprite.material.alphaTest = 0.1
return sprite, line
def ipv_animate(particles, velocities, particleColors, particleSizes,
paramRanges):
nTimesteps = particles.shape[0]
nParticles = particles.shape[1]
# determine quiver sizes and colors
veloQuiverSizes = np.zeros((nTimesteps, nParticles, 1))
for iTimestep in range(nTimesteps):
veloQuiverSizes[iTimestep, :,
0] = np.linalg.norm(velocities[iTimestep, :, :],
axis=1)
veloQuiverSizes = np.clip(veloQuiverSizes, 0, 2)
quiverColors = np.ones(
(nTimesteps, nParticles, 4)) * .8 #veloQuiverSizes/3
ipv.figure()
#bPlots = ipv.scatter( best[:, :, 0], best[:, :, 1], best[:, :, 2], marker = 'sphere', size=2, color = 'red' )
pPlots = ipv.scatter(particles[:, :, 0],
particles[:, :, 1],
particles[:, :, 2],
marker='sphere',
size=particleSizes,
color=particleColors)
#qvPlots = ipv.quiver( particles[:, :, 0], particles[:, :, 1], particles[:, :, 2], velocities[:, :, 0], velocities[:, :, 1], velocities[:, :, 2], size=veloQuiverSizes[:,:,0]*3, color=quiverColors)
ipv.animation_control([pPlots], interval=600)
ipv.xlim(paramRanges[0][1], paramRanges[0][2])
ipv.ylim(paramRanges[1][1], paramRanges[1][2])
ipv.zlim(paramRanges[2][1], paramRanges[2][2])
ipv.show()
def particleSimulate(num_particles,
box_size,
total_time,
time_step,
particle_radius,
grav=False,
save=False):
print("Running simulation...")
"""Run the simulation and extract the x,y,z coordinates to plot the particle's path"""
particle_list = initialize_particles(num_particles, box_size, time_step,
grav) #Generate starting points
x = np.zeros([total_time, num_particles, 1])
y = np.zeros([total_time, num_particles, 1])
z = np.zeros([total_time, num_particles, 1])
time = 0
#print(str(tot_eng(particle_list))) #Used to track the total energy of the particles
while time < total_time: #Loop through iterations of particle movements
#Check to see if bouncing occurs
isbounce(particle_list, particle_radius, time_step)
for i in range(len(particle_list)):
particle_list[i].pos_update(time_step) #Update position
x[time, i] = particle_list[i].pos[0]
y[time, i] = particle_list[i].pos[1]
z[time, i] = particle_list[i].pos[2]
time += 1
if (time / total_time) * 100 % 10 == 0:
print(str(time / total_time * 100) + "% complete")
"""Plot the results of all of the particle movements"""
colors = []
for i in range(num_particles):
colors.append(
[np.random.random(),
np.random.random(),
np.random.random()])
ipv.figure()
s = ipv.scatter(x, z, y, color=colors, size=7, marker="sphere")
ipv.animation_control(s, interval=1)
ipv.xlim(-1, box_size + 1)
ipv.ylim(-1, box_size + 1)
ipv.zlim(-1, box_size + 1)
ipv.style.axes_off()
ipv.show()
if save == True:
print("Saving the video of the simulation in the current directory...")
ipv.save('./particle_sim.html')
def visualize_sphere_grids(s2_grids,
params,
folder='grid',
colors=["#7b0001", "#ff0001", "#ff8db4"]):
fig = ipv.figure()
for _, layer_grids in s2_grids:
for i, (_, shell) in enumerate(layer_grids):
shell = shell.reshape(-1, 3).transpose(0, 1) # tensor([3, 4b^2]
x_axis = shell[0, :].cpu().numpy()
y_axis = shell[1, :].cpu().numpy()
z_axis = shell[2, :].cpu().numpy()
ipv.scatter(x_axis,
y_axis,
z_axis,
marker="sphere",
color=colors[i])
ipv.save("./imgs/{}/s2_sphere_sphere.html".format(folder))
def gaussian(N=1000, draw=True, show=True, seed=42, color=None, marker='sphere'):
"""Show N random gaussian distributed points using a scatter plot"""
import ipyvolume as ipv
rng = np.random.RandomState(seed)
x, y, z = rng.normal(size=(3, N))
if draw:
if color:
mesh = ipv.scatter(x, y, z, marker=marker, color=color)
else:
mesh = ipv.scatter(x, y, z, marker=marker)
if show:
#ipv.squarelim()
ipv.show()
return mesh
else:
return x, y, z
def stars(N=1000, radius=100000, thickness=3, seed=42, color=[255, 240, 240]):
import ipyvolume as ipv
rng = np.random.RandomState(seed)
x, y, z = rng.normal(size=(3, N))
r = np.sqrt(x**2 + y**2 + z**2)/(radius + thickness * radius * np.random.random(N))
x /= r
y /= r
z /= r
return ipv.scatter(x, y, z, texture=radial_sprite((64, 64), color), marker='square_2d', grow_limits=False, size=radius*0.7/100)
def test_ipyvolume():
import ipyvolume as ipv
s = 1 / 2**0.5
# 4 vertices for the tetrahedron
x = np.array([1., -1, 0, 0])
y = np.array([0, 0, 1., -1])
z = np.array([-s, -s, s, s])
# and 4 surfaces (triangles), where the number refer to the vertex index
triangles = [(0, 1, 2), (0, 1, 3), (0, 2, 3), (1, 3, 2)]
ipv.figure()
# draw the tetrahedron mesh
ipv.plot_trisurf(x, y, z, triangles=triangles, color='orange')
# mark the vertices
ipv.scatter(x, y, z, marker='sphere', color='blue')
# set limits and show
ipv.xyzlim(-2, 2)
ipv.show()
def ipv_plot_plydata(plydata, sample_frac=1, marker='circle_2d', **kwargs):
"""Plot vertices in a plydata object using ipyvolume."""
if sample_frac < 1:
xyz = random_sample(plydata['vertex'], plydata['vertex'].count,
sample_frac)
else:
xyz = dict(x=plydata['vertex']['x'],
y=plydata['vertex']['y'],
z=plydata['vertex']['z'])
fig = ipv.scatter(**xyz, marker=marker, **kwargs)
return fig
def show_patient_3d(src_dir, patient_name):
filepath = 'data/' + patient_name + '-mask.npy'
mask = np.load(filepath)
fig = ipv.figure()
colors = ['red', 'green', 'blue']
n_organs = 3
contours = {}
for organ_num in range(n_organs):
contours = mask2contours(mask[:,:,:,organ_num])
coord = np.argwhere(contours)
n_points = coord.shape[0]
x, y, z = np.random.normal(0,100,(3,n_points))
for i in range(n_points):
x[i] = coord[i,0]
y[i] = coord[i,1]
z[i] = coord[i,2]
if n_points>0:
scatter = ipv.scatter(x, y, z, size=1, marker='sphere', color=colors[organ_num])
ipv.show()
if src_dir != None:
prediction = np.load(src_dir + '/predictions/' + patient_name + '_prediction.npy')
prediction_thr = np.argmax(prediction, axis=-1)
fig = ipv.figure()
colors = ['magenta', 'lime', 'cyan']
n_organs = 3
contours = {}
for organ_num in range(n_organs):
mask = (prediction_thr==organ_num)
contours = mask2contours(mask)
coord = np.argwhere(contours)
n_points = coord.shape[0]
x, y, z = np.random.normal(0,100,(3,n_points))
for i in range(n_points):
x[i] = coord[i,0]
y[i] = coord[i,1]
z[i] = coord[i,2]
if n_points>0:
scatter = ipv.scatter(x, y, z, size=1, marker='sphere', color=colors[organ_num])
ipv.show()
def SC(L_x, L_y, L_z, Color='red'):
Ax = []
Ay = []
Az = []
for O in range(L_x):
#for each layer in the y direction.
for S in range(L_y):
#for each layer in the z direction.
for A in range(L_z):
Ax.append(float(O))
Ay.append(float(S))
Az.append(float(A))
#the x, y, z limits
ipv.xyzlim(0, 10) #define as a function highest of L_xyz
ipv.scatter(np.array(Ax),
np.array(Ay),
np.array(Az),
marker='sphere',
size=6.5,
color=Color)
ipv.show()