# Data¶

## A¶

**A** is a planar straight line graph of the capital letter A. We use it as
input to get a constrained Delaunay triangulation.

## BOX¶

**BOX** is a planar straight line graph of a double box. We use it as input to
get a constrained Delaunay triangulation.

The command:

```
triangulate(box, opts='pc')
```

creates the first mesh. (We need the pc because we didn’t surround our initial area with segments.)

The command:

```
box1 = get_data('box.1')
triangulate(box1, opts='rpa0.2')
```

refines the box.1 mesh with an area constraint of 0.2:

The command:

```
box2 = get_data('box.2')
triangulate(box2, opts='rpa0.05')
```

refines the box.2 mesh with an area constraint of 0.05, 1/4 of the previous maximum area:

The command “triangulate(box3, opts=’rpa0.0125’)” refines the box.3 mesh with an area constraint of 0.0125, 1/4 of the previous maximum area:

## DIAMOND_02_00009¶

**DIAMOND_02_00009** is another set of test data, for which we want the Voronoi
diagram.

## DOUBLE_HEX¶

**DOUBLE_HEX** describes a unit square with two hexagonal holes. 72 points are
listed on the outer boundary, and 12 on each of the holes. It is desired to
create a nice looking mesh of about 500 nodes, and no additional nodes on the
boundary segments.

Our first command:

```
triangle -p double_hex.poly
```

requests that we triangulate the current points:

Our second command:

```
triangle -pqY -a0.0015 double_hex.1.poly
```

requests that we triangulate the current points, adding new nodes as necessary to make a nice mesh, with no triangle being larger than 0.0015 in area, and with no points added on boundary segments. We end up with 525 nodes and 956 elements:

## DOUBLE_HEX2¶

**DOUBLE_HEX2** describes a unit square with two hexagonal holes. 36 points are
listed on the outer boundary, and 6 on each of the holes. It is desired to
create a nice looking mesh of about 235 elements, and no additional nodes on
the boundary segments.

Our first command:

```
triangle -p double_hex2.poly
```

requests that we triangulate the current points:

Our second command:

```
triangle -pqY -a0.0060 double_hex2.1.poly
```

requests that we triangulate the current points, adding new nodes as necessary to make a nice mesh, with no triangle being larger than 0.0060 in area, and with no points added on boundary segments. We end up with 141 nodes and 236 elements:

**DOUBLE_HEX3** describes a unit square with two hexagonal holes. 4 points are
listed on the outer boundary, and 6 on each of the holes. We want triangle to
triangulate this region.

Our command:

```
triangle -pq -a0.0015 double_hex3.poly
```

requests that we triangulate the region, adding points as necessary so that no triangle has an area greater than 0.0015.

## GREENLAND¶

**GREENLAND** is a triangulation of Greenland, using a graded (varying-size)
mesh of 33,343 nodes and 64,125 elements. The resulting image is essentially a
black blob the shape of Greenland. However, by modifying the code below, it is
possible to see interesting details of the mesh.

## LA¶

**LA** is a POLY file containing information representing soil layers. The data
includes points that are bounded by line segments defining the different
layers. The intent is that certain layers will be triangulated with smaller
area requirements.

## SPIRAL¶

**SPIRAL** is a node file containing points that form a spiral.

The command “triangle spiral” produces a Delaunay triangulation of the points, in the following node and element files:

The command “triangle -q spiral” produces a Delaunay triangulation with no angle smaller than 20 degrees (the default). This is done by adding points as necessary: in the following node and element files:

The command “triangle -q32.5 spiral” produces a Delaunay triangulation with no angle smaller than 32.5 degrees. This is done by adding points as necessary: in the following node and element files:

## SQUARE_CIRCLE_HOLE¶

**SQUARE_CIRCLE_HOLE** is a planar straight line graph of a square region with
an off center circular hole, and 826 points computed by a CVT calculation,
prepared by Hua Fei.