Note
Go to the end to download the full example code.
Edges in 3D
This tutorial show how to plot edges in 3D, using
oat_python.plot.trace_3d_for_edge(), which generates a trace for a single edge, andoat_python.plot.trace_3d_for_edges(), which generates a single trace for multiple edges.
Setup
import oat_python as oat
import plotly
import plotly.graph_objects as go
import numpy as np
We’ll use a cycle graph on N vertices as an example. Here we generate N evenly spaced points on a circle, which will be the vertices of the graph.
N = 60
# Generate N points in the xy-plane
points = oat.point_cloud.circle( n_points=N, )
# Add a z-coordinate to make the points 3D
points = np.hstack( (
points,
np.sin(np.linspace(0, 6*np.pi, N)).reshape(-1, 1))
)
points
array([[ 1.00000000e+00, 0.00000000e+00, 0.00000000e+00],
[ 9.94521895e-01, 1.04528463e-01, 3.14076712e-01],
[ 9.78147601e-01, 2.07911691e-01, 5.96367359e-01],
[ 9.51056516e-01, 3.09016994e-01, 8.18302776e-01],
[ 9.13545458e-01, 4.06736643e-01, 9.57422038e-01],
[ 8.66025404e-01, 5.00000000e-01, 9.99645611e-01],
[ 8.09016994e-01, 5.87785252e-01, 9.40700267e-01],
[ 7.43144825e-01, 6.69130606e-01, 7.86551556e-01],
[ 6.69130606e-01, 7.43144825e-01, 5.52800065e-01],
[ 5.87785252e-01, 8.09016994e-01, 2.63102564e-01],
[ 5.00000000e-01, 8.66025404e-01, -5.32221748e-02],
[ 4.06736643e-01, 9.13545458e-01, -3.64160575e-01],
[ 3.09016994e-01, 9.51056516e-01, -6.38244184e-01],
[ 2.07911691e-01, 9.78147601e-01, -8.47734428e-01],
[ 1.04528463e-01, 9.94521895e-01, -9.71429893e-01],
[ 6.12323400e-17, 1.00000000e+00, -9.96812007e-01],
[-1.04528463e-01, 9.94521895e-01, -9.21311978e-01],
[-2.07911691e-01, 9.78147601e-01, -7.52570770e-01],
[-3.09016994e-01, 9.51056516e-01, -5.07665800e-01],
[-4.06736643e-01, 9.13545458e-01, -2.11382624e-01],
[-5.00000000e-01, 8.66025404e-01, 1.06293486e-01],
[-5.87785252e-01, 8.09016994e-01, 4.13212186e-01],
[-6.69130606e-01, 7.43144825e-01, 6.78311836e-01],
[-7.43144825e-01, 6.69130606e-01, 8.74763085e-01],
[-8.09016994e-01, 5.87785252e-01, 9.82684125e-01],
[-8.66025404e-01, 5.00000000e-01, 9.91152831e-01],
[-9.13545458e-01, 4.06736643e-01, 8.99312130e-01],
[-9.51056516e-01, 3.09016994e-01, 7.16456740e-01],
[-9.78147601e-01, 2.07911691e-01, 4.61092501e-01],
[-9.94521895e-01, 1.04528463e-01, 1.59063496e-01],
[-1.00000000e+00, 1.22464680e-16, -1.59063496e-01],
[-9.94521895e-01, -1.04528463e-01, -4.61092501e-01],
[-9.78147601e-01, -2.07911691e-01, -7.16456740e-01],
[-9.51056516e-01, -3.09016994e-01, -8.99312130e-01],
[-9.13545458e-01, -4.06736643e-01, -9.91152831e-01],
[-8.66025404e-01, -5.00000000e-01, -9.82684125e-01],
[-8.09016994e-01, -5.87785252e-01, -8.74763085e-01],
[-7.43144825e-01, -6.69130606e-01, -6.78311836e-01],
[-6.69130606e-01, -7.43144825e-01, -4.13212186e-01],
[-5.87785252e-01, -8.09016994e-01, -1.06293486e-01],
[-5.00000000e-01, -8.66025404e-01, 2.11382624e-01],
[-4.06736643e-01, -9.13545458e-01, 5.07665800e-01],
[-3.09016994e-01, -9.51056516e-01, 7.52570770e-01],
[-2.07911691e-01, -9.78147601e-01, 9.21311978e-01],
[-1.04528463e-01, -9.94521895e-01, 9.96812007e-01],
[-1.83697020e-16, -1.00000000e+00, 9.71429893e-01],
[ 1.04528463e-01, -9.94521895e-01, 8.47734428e-01],
[ 2.07911691e-01, -9.78147601e-01, 6.38244184e-01],
[ 3.09016994e-01, -9.51056516e-01, 3.64160575e-01],
[ 4.06736643e-01, -9.13545458e-01, 5.32221748e-02],
[ 5.00000000e-01, -8.66025404e-01, -2.63102564e-01],
[ 5.87785252e-01, -8.09016994e-01, -5.52800065e-01],
[ 6.69130606e-01, -7.43144825e-01, -7.86551556e-01],
[ 7.43144825e-01, -6.69130606e-01, -9.40700267e-01],
[ 8.09016994e-01, -5.87785252e-01, -9.99645611e-01],
[ 8.66025404e-01, -5.00000000e-01, -9.57422038e-01],
[ 9.13545458e-01, -4.06736643e-01, -8.18302776e-01],
[ 9.51056516e-01, -3.09016994e-01, -5.96367359e-01],
[ 9.78147601e-01, -2.07911691e-01, -3.14076712e-01],
[ 9.94521895e-01, -1.04528463e-01, -7.34788079e-16]])
Color scales
Plotly has no native functionality to plot multiple edges on a colorscale in a single trace. If you’d like to plot multiple edges on a colorscale, you can create a separate trace for each edge, and assign a color to each edge individually. Here’s an example.
Get a Plotly colorscale object, which maps values in [0,1] to RGB colors.
For a full list of Plotly colorscales, see: https://plotly.com/python/builtin-colorscales/
For another example of mapping scalar values to colors, see Color Mapping.
colorscale = plotly.colors.get_colorscale('Rainbow')
Convert N evenly spaced values in [0,1] to colors using the colorscale. We’ll use these colors for the edges of a cycle graph.
colors = plotly.colors.sample_colorscale(
colorscale,
np.linspace(0, 1, N)
)
colors
['rgb(150, 0, 90)', 'rgb(130, 0, 105)', 'rgb(109, 0, 120)', 'rgb(89, 0, 135)', 'rgb(69, 0, 150)', 'rgb(48, 0, 165)', 'rgb(28, 0, 179)', 'rgb(8, 0, 194)', 'rgb(0, 2, 205)', 'rgb(0, 6, 212)', 'rgb(0, 9, 220)', 'rgb(0, 12, 227)', 'rgb(0, 16, 234)', 'rgb(0, 19, 242)', 'rgb(0, 22, 249)', 'rgb(0, 29, 255)', 'rgb(0, 47, 255)', 'rgb(0, 64, 255)', 'rgb(0, 81, 255)', 'rgb(0, 98, 255)', 'rgb(0, 115, 255)', 'rgb(0, 133, 255)', 'rgb(0, 150, 255)', 'rgb(5, 164, 243)', 'rgb(11, 178, 228)', 'rgb(17, 192, 214)', 'rgb(23, 206, 200)', 'rgb(29, 220, 186)', 'rgb(35, 234, 171)', 'rgb(41, 248, 157)', 'rgb(51, 255, 140)', 'rgb(66, 255, 119)', 'rgb(80, 255, 99)', 'rgb(95, 255, 79)', 'rgb(109, 255, 58)', 'rgb(124, 255, 38)', 'rgb(138, 255, 18)', 'rgb(153, 255, 0)', 'rgb(167, 252, 0)', 'rgb(181, 249, 0)', 'rgb(195, 246, 0)', 'rgb(209, 243, 0)', 'rgb(223, 240, 0)', 'rgb(237, 238, 0)', 'rgb(251, 235, 0)', 'rgb(255, 221, 0)', 'rgb(255, 205, 0)', 'rgb(255, 188, 0)', 'rgb(255, 171, 0)', 'rgb(255, 155, 0)', 'rgb(255, 138, 0)', 'rgb(255, 121, 0)', 'rgb(255, 105, 0)', 'rgb(255, 90, 0)', 'rgb(255, 75, 0)', 'rgb(255, 60, 0)', 'rgb(255, 45, 0)', 'rgb(255, 30, 0)', 'rgb(255, 15, 0)', 'rgb(255, 0, 0)']
Plot the edges of the graph, using oat_python.plot.trace_2d_for_edge().
# Initialize a list to hold the traces
traces = []
# Make a trace for the vertices, to help mark start and end points
vertex_trace = oat.plot.trace_3d_for_vertices(
points = points,
marker = dict(
color = 'white',
size = 5,
),
showlegend = True,
name = "Vertex Trace",
)
traces.append(vertex_trace)
# Make one trace per edge, each with a different color. For an explanation of
# the keyword arguments, see the docstring for :func:`oat_python.plot.trace_2d_for_edge`.
for p in range(N):
edge_trace = oat.plot.trace_3d_for_edge(
points = points,
edge = [p, (p+1) % N],
line = dict(
color = colors[p],
width = 6,
),
name = f"Trace for edge {p}",
)
traces.append(edge_trace)
# Create a figure with all traces
fig = plotly.graph_objects.Figure(traces)
# Set the title, and make the aspect ratio equal
fig.update_layout(
title="Using a colorscale by placing edges in separate traces",
scene = dict(
aspectratio=go.layout.scene.Aspectratio(x=1, y=1, z=1),
)
)
# Show the figure
fig
See the Styling in 3D for tips on styling in 3D.
# Apply a dark template
fig.update_layout(template="plotly_dark")
fig
Grouping legend entries
Each trace appears separately in the legend by default. However,
You can group multiple traces together in the legend using keyword arguments. See the Legend section of Triangles in 3D for examples.
If you do not need to assign different colors to different edges, then you can plot multiple edges in a single trace using
oat_python.plot.trace_2d_for_edges(). See the next section for an example.
Grouping multiple edges in a single trace (may improve performance)
If you do not need to assign different colors to different edges, then
you can plot multiple edges in a single trace using oat_python.plot.trace_2d_for_edges().
Here’s an example.
# Make a trace for the vertices, to help mark start and end points
vertex_trace = oat.plot.trace_3d_for_vertices(
points = points,
marker = dict(
color = 'white',
size = 5,
),
showlegend = True,
name = "Vertex Trace",
)
Make a single trace for all edges, with the same color for all edges.
For an explanation of the keyword arguments, see the docstring for oat_python.plot.trace_2d_for_edges().
edge_trace = oat.plot.trace_3d_for_edges(
points = points,
edges = [ [p, (p+1) % N] for p in range(N) ],
line = dict(
color = "white",
width = 3,
),
showlegend = True,
name = f"Edge Trace",
)
# Create a figure with the trace
fig = plotly.graph_objects.Figure([ edge_trace, vertex_trace ])
# Set the title, make the aspect ratio equal, and apply a dark template
fig.update_layout(
title="Multiple edges in a single trace",
template="plotly_dark",
scene = dict(
aspectratio=go.layout.scene.Aspectratio(x=1, y=1, z=1),
)
)
# Show the figure
fig
Total running time of the script: (0 minutes 0.090 seconds)