Points Marker Style

[banesullivan]

• [ ] Can we edit how the points are displayed?
• [x] Could we use a circle instead of a square? Edit: We can render points as spheres - which appear as 2D if lighting is false
• [ ] Can we add an outline to all points similar to the matplotlib figure below

matplotlib figure of some hard data points and an interpolated grid

vtki figure of the same data (note how truth points are barely visible)

[banesullivan]

AFAIK, this isn't possible

[ramaguirre]

Hello @banesullivan

yes to all three, this can be done by updating the Fragment Shader. I was referred to this post (https://discourse.vtk.org/t/different-shaped-vertex-points/10136).

Example below, combining Maximum Intensity Projection with circular points, with and without boders. It works by discarding pixels located outside a radius of 0.5, and optionally assigning ambientColor = diffuseColor = (0,0,0) for a radius of 0.3 for a black outline.

There is a lot of potential in Shader replacement.

import vtk
import numpy as np
import pyvista as pv
from pyvista import examples
import matplotlib.pyplot as plt

####################################################################
######## MAXIMUM INTENSITY PROJECTION (MIP) FOR POINT CLOUDS #######
########         Ramón Aguirre M. 2024                       #######
####################################################################

#Donwload Knee dataset
vol = examples.download_knee_full()

# Get a Sample of vertices from the dataset
sample_n=100000
randix=np.random.randint(0,vol.n_points,sample_n)
pts=vol.extract_points(randix,adjacent_cells=False,include_cells=False).cast_to_pointset()

#Start pyvista plotter
p=pv.Plotter(shape=(1,3))
p.enable_parallel_projection()

#Define actors parameters
cmap='jet'
clim=[0,100]
opacity=1
rpas=False
lighting=False

#Plot normally for comparison
p.subplot(0,0)
#add point cloud actor to render window
point_actor=p.add_mesh(pts.flip_z(),cmap=cmap,opacity=opacity,clim=clim,
                       render_points_as_spheres=rpas,lighting=lighting,point_size=10)

p.add_title('Knee dataset\nNormal Projection\n(point cloud)',font_size=8)

#Implement MIP
p.subplot(0,1)
#add point cloud actor to render window
point_actor=p.add_mesh(pts.flip_z(),cmap=cmap,opacity=opacity,clim=clim,
                       render_points_as_spheres=rpas,lighting=lighting,point_size=10)

# Maximimum Intensity Projection MIP does not work with opacity <1
# unless Depth Peeling is enabled
if point_actor.prop.GetOpacityMinValue()<1:
    p.enable_depth_peeling(number_of_peels=None,occlusion_ratio=0)

# Get the Vertex Shader Program
# it runs once for each vertex every time the render is updated
# MAIN TRICK: we reset the z screen coordinates so vertices
# with higher values are closer to the screen (Cowan, 2014)
property=point_actor.GetShaderProperty()

max_scalar=np.nanmax(point_actor.mapper.dataset.active_scalars)
min_scalar=np.nanmin(point_actor.mapper.dataset.active_scalars)

property.AddShaderReplacement(
    vtk.vtkShader.Vertex,
    "//VTK::LineWidthGLES30::Impl", # We replace this text in the shader program
    True, # before the standard replacements
    f'''
    gl_Position.z = -(abs(colorTCoord[0])-{min_scalar})/({max_scalar}-{min_scalar});  //We add this line to set vertices with highest scalar values closer to the screen
      //VTK::LineWidthGLES30::Impl", //Keeping the original text
    ''',
    False # only do it once
)

property.AddShaderReplacement(
    vtk.vtkShader.Fragment,  # in the fragment shader
     "//VTK::Color::Impl", # replace the color block
     True, # before the standard replacements
     '''
     //VTK::Color::Impl // we still want the default calc
     //error
     float dist = length(vec2(gl_PointCoord.x - 0.5, gl_PointCoord.y - 0.5)); //Distance to centroid
     if (dist > 0.3) { ambientColor = vec3(0,0,0); //Set border colors to black
     diffuseColor = vec3(0,0,0);}   //Set border colors to black
     if (dist > 0.5) {discard;}''',  #but we add this: remove points outside of radius
     False # only do it once
)

p.add_title('Knee dataset\nMaximum Intensity Projection\n(point cloud)',font_size=8)

p.subplot(0,2)
#add point cloud actor to render window
point_actor=p.add_mesh(pts.flip_z(),cmap=cmap,opacity=opacity,clim=clim,
                       render_points_as_spheres=rpas,lighting=lighting,point_size=10)

# Maximimum Intensity Projection MIP does not work with opacity <1
# unless Depth Peeling is enabled
if point_actor.prop.GetOpacityMinValue()<1:
    p.enable_depth_peeling(number_of_peels=None,occlusion_ratio=0)

# Get the Vertex Shader Program
# it runs once for each vertex every time the render is updated
# MAIN TRICK: we reset the z screen coordinates so vertices
# with higher values are closer to the screen (Cowan, 2014)
property=point_actor.GetShaderProperty()

max_scalar=np.nanmax(point_actor.mapper.dataset.active_scalars)
min_scalar=np.nanmin(point_actor.mapper.dataset.active_scalars)

#Implement MIP
property.AddShaderReplacement(
    vtk.vtkShader.Vertex,
    "//VTK::LineWidthGLES30::Impl", # We replace this text in the shader program
    True, # before the standard replacements
    f'''
    gl_Position.z = -(abs(colorTCoord[0])-{min_scalar})/({max_scalar}-{min_scalar});  //We add this line to set vertices with highest scalar values closer to the screen
      //VTK::LineWidthGLES30::Impl", //Keeping the original text
    ''',
    False # only do it once
)

#Make markers circular
property.AddShaderReplacement(
    vtk.vtkShader.Fragment,  # in the fragment shader
     "//VTK::Color::Impl", # replace the color block
     True, # before the standard replacements
     '''
     //VTK::Color::Impl // we still want the default calc
     //error
     float dist = length(vec2(gl_PointCoord.x - 0.5, gl_PointCoord.y - 0.5)); //Distance to centroid
     if (dist > 0.5) {discard;}''', #but we add this: remove points outside of radius
     False # only do it once
)

p.add_title('Knee dataset\nMaximum Intensity Projection\n(point cloud)',font_size=8)

# Final settings and render

p.link_views()

#p.show(jupyter_backend) # MIP doas not work with 'html' backend
p.show() 

#Reference:
# 2014: Cowan, E.J., ‘X-ray Plunge Projection’— Understanding Structural Geology from Grade Data, AusIMM Monograph 30: 
# Mineral Resource and Ore Reserve Estimation — The AusIMM Guide to Good Practice, second edition, 207-220