Shuenhoy/vtk-m
1
0
Fork 0
mirror of https://gitlab.kitware.com/vtk/vtk-m synced 2024-10-05 01:49:02 +00:00
Code Issues 2 Actions Packages Projects Releases Wiki Activity

Merge topic 'document-merge-data-sets'

Browse Source 
45ad2fb50 Add documentation of MergeDataSets filter to user's guide

Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !3161
This commit is contained in:
Kenneth Moreland 2023-12-02 00:17:05 +00:00 committed by Kitware Robot
commit 5d0bf1d866
2 changed files with 29 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
docs/users-guide/provided-filters.rst
View File

@ -1004,6 +1004,19 @@ The :class:`vtkm::filter::multi_block::AmrArrays` filter determines the hierarch
This is why they use the prefix of "vtk" instead of "vtkm".
Likewise, the flags used for ``vtkGhostType`` are compatible with VTK.
Merge Data Sets
------------------------------
.. index::
double: merge data sets; filter
A :class:`vtkm::cont::PartitionedDataSet` can often be treated the same as a :class:`vtkm::cont::DataSet` as both can be passed to a filter's `Execute` method.
However, it is sometimes important to have all the data contained in a single ``DataSet``.
The :class:`vtkm::filter::multi_block::MergeDataSets` filter can do just that to the partitions of a `vtkm::cont::PartitionedDataSet`.
.. doxygenclass:: vtkm::filter::multi_block::MergeDataSets
:members:
Resampling
==============================

18
vtkm/filter/multi_block/MergeDataSets.h
View File

@ -20,15 +20,29 @@ namespace filter
{
namespace multi_block
{
/// \brief Merging multiple data sets into one data set.
/// @brief Merging multiple data sets into one data set.
///
/// This filter merges multiple data sets into one data set. We assume that the input data sets
/// have the same coordinate system. If there are missing fields in a specific data set,
// the filter uses the InvalidValue specified by the user to fill in the associated position of the field array.
/// the filter uses the InvalidValue specified by the user to fill in the associated position
/// of the field array.
///
/// `MergeDataSets` is used by passing a `vtkm::cont::PartitionedDataSet` to its `Execute()`
/// method. The `Execute()` will return a `vtkm::cont::PartitionedDataSet` because that is
/// the common interface for all filters. However, the `vtkm::cont::PartitionedDataSet` will
/// have one partition that is all the blocks merged together.
class VTKM_FILTER_MULTI_BLOCK_EXPORT MergeDataSets : public vtkm::filter::Filter
{
public:
/// @brief Specify the value to use where field values are missing.
///
/// One issue when merging blocks in a paritioned dataset is that the blocks/partitions
/// may have different fields. That is, one partition might not have all the fields of
/// another partition. When these partitions are merged together, the values for this
/// missing field must be set to something. They will be set to this value, which defaults
/// to NaN.
void SetInvalidValue(vtkm::Float64 invalidValue) { this->InvalidValue = invalidValue; };
/// @copydoc SetInvalidValue
vtkm::Float64 GetInvalidValue() { return this->InvalidValue; }
private: