tutorial

Author: tomvanmele

docs/examples/intersections.py

@@ -2,7 +2,7 @@
 import compas
 import compas_libigl as igl
 
-from compas.geometry import Point, Line, Translation
+from compas.geometry import Point, Line
 from compas.datastructures import Mesh
 from compas.colors import Color
 
@@ -14,12 +14,6 @@
 
 mesh = Mesh.from_obj(compas.get("tubemesh.obj"))
 
-z = mesh.vertices_attribute("z")
-zmin = min(z)
-
-T = Translation.from_vector([0, 0, -zmin])
-mesh.transform(T)
-
 trimesh = mesh.copy()
 trimesh.quads_to_triangles()
 

docs/tutorial.rst

@@ -2,12 +2,115 @@
 Tutorial
 ********************************************************************************
 
-* Similar input parameters as hte original code
-* Meshes represented by a list of vertices and a list of triangles
-* Both Python lists and Numpy arrays are supported
-* Most functions are implemented as pluggables for plugins of COMPAS core
-* Therefore, the code can be used from compas_libigl or from compas directly
-* Results can easily be visualized using the COMPAS Viewer
-* Since the binding is generated with PyBind11 and wraps C++ code, it cannot be used in Rhino/Grasshopper directly
-* However it can be used there through RPC
+.. rst-class:: lead
 
+:mod:`compas_libigl` provides bindings for the libigl library.
+It doesn't cover the entire library, but only for specific functions.
+Currently, the following functions are supported:
+
+* :func:`compas_libigl.intersection_ray_mesh`
+* :func:`compas_libigl.intersection_rays_mesh`
+* :func:`compas_libigl.trimesh_boundaries`
+* :func:`compas_libigl.trimesh_gaussian_curvature`
+* :func:`compas_libigl.trimesh_principal_curvature`
+* :func:`compas_libigl.trimesh_geodistance`
+* :func:`compas_libigl.trimesh_isolines`
+* :func:`compas_libigl.trimesh_massmatrix`
+* :func:`compas_libigl.trimesh_harmonic`
+* :func:`compas_libigl.trimesh_lscm`
+* :func:`compas_libigl.trimesh_remesh_along_isoline`
+* :func:`compas_libigl.quadmesh_planarize`
+
+
+Input/Output
+============
+
+The function signatures of the bindings are similar to the original libigl functions.
+Meshes are represented by a tuple containing a list/array of vertices and a list/array of faces.
+Most functions require the input mesh to be a triangle mesh.
+
+.. code-block:: python
+
+    import compas
+    import compas_libigl
+    from compas.datastructures import Mesh
+
+    mesh = Mesh.from_obj(compas.get("tubemesh.obj"))
+    mesh.quads_to_triangles()
+
+    V, F = mesh.to_vertices_and_faces()
+
+    source = trimesh.vertex_sample(size=1)[0]
+    distance = compas_libigl.trimesh_geodistance(
+        (V, F),
+        source,
+        method="heat",
+    )
+
+
+Both Python lists and Numpy arrays are supported.
+
+.. code-block:: python
+
+    import numpy
+    import compas
+    import compas_libigl
+    from compas.datastructures import Mesh
+
+    mesh = Mesh.from_obj(compas.get("tubemesh.obj"))
+    mesh.quads_to_triangles()
+
+    vertices, faces = mesh.to_vertices_and_faces()
+    V = numpy.array(vertices, dtype=float)
+    F = numpy.array(faces, dtype=int)
+
+    source = trimesh.vertex_sample(size=1)[0]
+    distance = compas_libigl.trimesh_geodistance(
+        (V, F),
+        source,
+        method="heat",
+    )
+
+
+Pluggables
+==========
+
+
+Visualisation
+=============
+
+
+Working in Rhino/Grasshopper
+============================
+
+The bindings are generated with PyBind11 and wrap the C++ code of libigl.
+Therefore, the bindings are not compatible with IronPython and cannot be used in Rhino/Grasshopper directly.
+However, they can be used in Rhino/Grasshopper through RPC.
+
+.. code-block:: python
+
+    import compas
+    from compas.rpc import Proxy
+    from compas.datastructures import Mesh
+    from compas.datastructures import mesh_flatness
+    from compas.colors import Color, ColorMap
+    from compas.artists import Artist
+
+    compas_libigl = Proxy('compas_libigl')
+
+    mesh = Mesh.from_obj(compas.get("tubemesh.obj"))
+    
+    V, F = mesh.to_vertices_and_faces()
+    V2 = compas_libigl.quadmesh_planarize((V, F), 100, 0.005)
+
+    mesh = Mesh.from_vertices_and_faces(V2, F)
+    dev = mesh_flatness(mesh, maxdev=TOL)
+    cmap = ColorMap.from_two_colors(Color.white(), Color.blue())
+
+    facecolor={
+        face: (cmap(dev[face]) if dev[face] <= 1.0 else Color.red())
+        for face in mesh.faces()
+    }
+
+    artist = Artist(mesh, layer="libigl::quadmesh_planarize")
+    artist.draw(facecolor=facecolor, disjoint=True)