ParaView can work with many different kinds of data when generating visualizations. This data can take any of the following forms:

Regular Grid

This type of data includes 2D images and 3D volume datasets. Data points are evenly spaced in each dimension of the grid. Examples include data from MRIs and CAT scans.

Rectilinear Grid

These are similar to Regular Grids in that the data is arranged along orthogonal axes, but the data need not be evenly spaced along all axes.

Structured Grid

These data points in these grids are regularly spaced but are not aligned to orthogonal axes. Examples include subdivided irregular volumes from fluid flow or heat transfer analysis.

Unstructured Grid

Here, data points are neither evenly spaced nor orthogonally aligned. Finite element analysis data fits this category.

Polygons

All data points in these datasets are vertices of polygons or line-strings. All data points in these datasets are vertices of polygons or line-strings. No grid structure is present. Such data might come from CAD models.

Unstructured Points

This category includes all forms of data that do not fit the categories above.

Each data source can contain many data points, and each point must have at least a geometric location. Points may also have additional data that can be used to generate visualizations. Such data could include:

Scalars

Single values that might represent a measurement such as density, temperature or pressure.

Vectors

Multiple values that store a multi-dimensional direction and magnitude, possibly representing values such as velocity or momentum.

Normals

A special case of Vectors, this data has the same dimension as the geometry and is "normalized" so that each vector has length 1.0. This data is typically found in polygonal models and is used to smoothly shade surfaces.

Texture Coordinates

Vectors used to correlate a data point to a location in some texture space, often of a different dimension than the source geometry. These coordinates are commonly used to select data values from another dataset, such as pixel colors from an image that is mapped onto a polygonal surface.

Tensors

A generalization of scalars and vectors, with rank 0 a scalar, rank 1 a vector, rank 2 a 2D matrix, etc.