This chapters discusses how to create a task graph dynamically using asynchronous tasks, which is extremely beneficial for workloads that want to (1) explore task graph parallelism out of dynamic control flow or (2) overlap task graph creation time with individual task execution time. We recommend that you first read Asynchronous Tasking before digesting this chapter.
When the construct-and-run model of a task graph is not possible in your application, you can use tf::A, B, C, and D, where A runs before B and C and D runs after B and C:
The task spawns asynchronous tasks to perform parallel reduction over init and the transformed elements in the range [first, last). The reduced result is store in init. This method is equivalent to the parallel execution of the following loop:
for(auto itr=first; itr!=last; itr++) { init = bop(init, uop(*itr)); +}
Iterators are templated to enable stateful range using std::
Please refer to Parallel Reduction for details.
+constructs an STL-styled parallel transform-reduce task
+| Template parameters | |
|---|---|
| B1 | +first beginning iterator type | +
| E1 | +first ending iterator type | +
| B2 | +second beginning iterator type | +
| T | +result type | +
| BOP_R | +binary reducer type | +
| BOP_T | +binary transformion type | +
| P | +partitioner type (default tf:: |
+
| Parameters | |
| first1 | ++ |
| last1 | ++ |
| first2 | ++ |
| init | +initial value of the reduction and the storage for the reduced result | +
| bop_r | +binary operator that will be applied in unspecified order to the results of bop_t |
+
| bop_t | +binary operator that will be applied to transform each element in the range to the result type | +
| part | +partitioning algorithm to schedule parallel iterations | +
| Returns | +a tf:: |
+
The task spawns asynchronous tasks to perform parallel reduction over init and the transformed elements in the range [first, last). The reduced result is store in init. This method is equivalent to the parallel execution of the following loop:
for(auto itr1=first1, itr2=first2; itr1!=last1; itr1++, itr2++) { + init = bop_r(init, bop_t(*itr1, *itr2)); }
Iterators are templated to enable stateful range using std::
Please refer to Parallel Reduction for details.
Taskflow 3.7.0 is the newest developing line to new features and improvements we continue to support. It is also where this documentation is generated. Many things are considered experimental and may change or break from time to time. While it may be difficult to be keep all things consistent when introducing new features, we continue to try our best to ensure backward compatibility.
To download the newest version of Taskflow, please clone the master branch from Taskflow's GitHub.
To use Taskflow v3.7.0, you need a compiler that supports C++17:
Taskflow works on Linux, Windows, and Mac OS X.
Please do not hesitate to contact Dr. Tsung-Wei Huang if you intend to collaborate with us on using Taskflow in your scientific computing projects.
Taskflow 3.7.0 is the newest developing line to new features and improvements we continue to support. It is also where this documentation is generated. Many things are considered experimental and may change or break from time to time. While it may be difficult to be keep all things consistent when introducing new features, we continue to try our best to ensure backward compatibility.
To download the newest version of Taskflow, please clone the master branch from Taskflow's GitHub.
To use Taskflow v3.7.0, you need a compiler that supports C++17:
Taskflow works on Linux, Windows, and Mac OS X.
Please do not hesitate to contact Dr. Tsung-Wei Huang if you intend to collaborate with us on using Taskflow in your scientific computing projects.