Line data Source code
1 1 : // Distributed under the MIT License.
2 : // See LICENSE.txt for details.
3 :
4 : /// \file
5 : /// Defines all group definitions
6 :
7 : #pragma once
8 :
9 : /*!
10 : * \defgroup ActionsGroup Actions
11 : * \brief A collection of steps used in algorithms.
12 : */
13 :
14 : /*!
15 : * \defgroup AnalyticDataGroup Analytic Data
16 : * \brief Analytic data used to specify (for example) initial data to the
17 : * equations implemented in \ref EvolutionSystemsGroup.
18 : */
19 :
20 : /*!
21 : * \defgroup AnalyticSolutionsGroup Analytic Solutions
22 : * \brief Analytic solutions to the equations implemented in \ref
23 : * EvolutionSystemsGroup and \ref EllipticSystemsGroup.
24 : */
25 :
26 : /*!
27 : * \defgroup BoundaryConditionsGroup Boundary Conditions
28 : * A collection of boundary conditions used for evolutions.
29 : */
30 :
31 : /*!
32 : * \defgroup CharmExtensionsGroup Charm++ Extensions
33 : * \brief Classes and functions used to make Charm++ easier and safer to use.
34 : */
35 :
36 : /*!
37 : * \defgroup ComputationalDomainGroup Computational Domain
38 : * \brief The building blocks used to describe the computational domain.
39 : *
40 : * ### Description
41 : * The VolumeDim-dimensional computational Domain is constructed from a set of
42 : * non-overlapping Block%s. Each Block is a distorted VolumeDim-dimensional
43 : * hypercube. Each codimension-1 boundary of a Block is either part of the
44 : * external boundary of the computational domain, or is identical to a boundary
45 : * of one other Block. Each Block is subdivided into one or more Element%s
46 : * that may be changed dynamically if AMR is enabled.
47 : */
48 :
49 : /*!
50 : * \defgroup ConservativeGroup Conservative System Evolution
51 : * \brief Contains generic functions used for evolving conservative
52 : * systems.
53 : */
54 :
55 : /*!
56 : * \defgroup ConstantExpressionsGroup Constant Expressions
57 : * \brief Contains an assortment of constexpr functions
58 : *
59 : * ### Description
60 : * Contains an assortment of constexpr functions that are useful for
61 : * metaprogramming, or efficient mathematical computations, such as
62 : * exponentiating to an integer power, where the power is known at compile
63 : * time.
64 : */
65 :
66 : /*!
67 : * \defgroup ControlSystemGroup Control System
68 : * \brief Contains control system elements
69 : *
70 : * The control system manages the time-dependent mapping between frames, such as
71 : * the fixed computational frame (grid frame) and the inertial frame. The
72 : * time-dependent parameters of the mapping are adjusted by a feedback control
73 : * system in order to follow the dynamical evolution of objects such as horizons
74 : * of black holes or surfaces of neutron stars. For example, in binary black
75 : * hole simulations the map is typically a composition of maps that include
76 : * translation, rotation, scaling, shape, etc.
77 : * Each map under the governance of the control system has an associated
78 : * time-dependent map parameter \f$\lambda(t)\f$ that is a piecewise Nth order
79 : * polynomial. At discrete times (called reset times), the control system resets
80 : * the Nth time derivative of \f$\lambda(t)\f$ to a new constant value, in order
81 : * to minimize an error function \f$Q(t)\f$ that is specific to each map. At
82 : * each reset time, the Nth derivative of \f$\lambda(t)\f$ is set to a function
83 : * \f$U(t)\f$, called the control signal, that is determined by \f$Q(t)\f$ and
84 : * its time derivatives and time integral. Note that \f$\lambda(t)\f$,
85 : * \f$U(t)\f$, and \f$Q(t)\f$ can be vectors.
86 : *
87 : * The key components of the control system are:
88 : * - FunctionsOfTime: each map has an associated FunctionOfTime that represents
89 : * the map parameter \f$\lambda(t)\f$ and relevant time derivatives.
90 : * - ControlError: each map has an associated ControlError that computes
91 : * the error, \f$Q(t)\f$. Note that for each map, \f$Q(t)\f$ is defined to
92 : * follow the convention that \f$dQ = -d \lambda\f$ as \f$Q \rightarrow 0\f$.
93 : * - Averager: an averager can be used to average out the noise in the 'raw'
94 : * \f$Q(t)\f$ returned by the ControlError.
95 : * - Controller: the map controller computes the control signal \f$U(t)\f$ from
96 : * \f$Q(t)\f$ and its time integral and time derivatives.
97 : * The control is accomplished by setting the Nth derivative of
98 : * \f$\lambda(t)\f$ to \f$U(t)\f$. Two common controllers are PID
99 : * (proportional/integral/derivative)
100 : * \f[U(t) = a_{0}\int_{t_{0}}^{t} Q(t') dt'+a_{1}Q(t)+a_{2}\frac{dQ}{dt}\f]
101 : * or
102 : * PND (proportional/N derivatives)
103 : * \f[ U(t) = \sum_{k=0}^{N} a_{k} \frac{d^kQ}{dt^k} \f]
104 : * The coefficients \f$ a_{k} \f$ in the computation of \f$U(t)\f$ are chosen
105 : * at each time such that the error \f$Q(t)\f$ will be critically damped
106 : * on a timescale of \f$\tau\f$ (the damping time),
107 : * i.e. \f$Q(t) \propto e^{-t/\tau}\f$.
108 : * - TimescaleTuner: each map has a TimescaleTuner that dynamically adjusts
109 : * the damping timescale \f$\tau\f$ appropriately to keep the error \f$Q(t)\f$
110 : * within some specified error bounds. Note that the reset time interval,
111 : * \f$\Delta t\f$, is a constant fraction of this damping timescale,
112 : * i.e. \f$\Delta t = \alpha \tau\f$ (empirically, we have found
113 : * \f$\alpha=0.3\f$ to be a good choice).
114 : *
115 : *
116 : * For additional details describing our control system approach, see
117 : * \cite Hemberger2012jz.
118 : */
119 :
120 : /*!
121 : * \defgroup CoordinateMapsGroup Coordinate Maps
122 : * \brief Functions for mapping coordinates between different frames
123 : *
124 : * Coordinate maps provide the maps themselves, the inverse maps, along
125 : * with the Jacobian and inverse Jacobian of the maps.
126 : */
127 :
128 : /*!
129 : * \defgroup CoordMapsTimeDependentGroup Coordinate Maps, Time-dependent
130 : * \brief Functions for mapping time-dependent coordinates between different
131 : * frames
132 : *
133 : * Coordinate maps provide the maps themselves, the inverse maps, the Jacobian
134 : * and inverse Jacobian of the maps, and the frame velocity (time derivative of
135 : * the map)
136 : */
137 :
138 : /*!
139 : * \defgroup DataBoxGroup DataBox
140 : * \brief Documentation, functions, metafunctions, and classes necessary for
141 : * using DataBox
142 : *
143 : * DataBox is a heterogeneous compile-time associative container with lazy
144 : * evaluation of functions. DataBox can not only store data, but can also store
145 : * functions that depend on other data inside the DataBox. The functions will be
146 : * evaluated when the data they return is requested. The result is cached, and
147 : * if a dependency of the function is modified the cache is invalidated.
148 : *
149 : * #### Simple and Compute Tags and Their Items
150 : *
151 : * The compile-time keys are `struct`s called tags, while the values are called
152 : * items. Tags are quite minimal, containing only the information necessary to
153 : * store the data and evaluate functions. There are two different types of tags
154 : * that a DataBox can hold: simple tags and compute tags. Simple tags are for
155 : * data that is inserted into the DataBox at the time of creation, while compute
156 : * tags are for data that will be computed from a function when the compute item
157 : * is retrieved. If a compute item is never retrieved from the DataBox then it
158 : * is never evaluated.
159 : *
160 : * Simple tags must have a member type alias `type` that is the type of the data
161 : * to be stored and a `static std::string name()` method that returns the name
162 : * of the tag. Simple tags must inherit from `db::SimpleTag`.
163 : *
164 : * Compute tags must also have a `static std::string name()` method that returns
165 : * the name of the tag, but they cannot have a `type` type alias. Instead,
166 : * compute tags must have a static member function or static member function
167 : * pointer named `function`. `function` can be a function template if necessary.
168 : * The `function` must take all its arguments by `const` reference. The
169 : * arguments to the function are retrieved using tags from the DataBox that the
170 : * compute tag is in. The tags for the arguments are set in the member type
171 : * alias `argument_tags`, which must be a `tmpl::list` of the tags corresponding
172 : * to each argument. Note that the order of the tags in the `argument_list` is
173 : * the order that they will be passed to the function. Compute tags must inherit
174 : * from `db::ComputeTag`.
175 : *
176 : * Here is an example of a simple tag:
177 : *
178 : * \snippet Test_DataBox.cpp databox_tag_example
179 : *
180 : * and an example of a compute tag with a function pointer:
181 : *
182 : * \snippet Test_DataBox.cpp databox_mutating_compute_item_tag
183 : *
184 : * If the compute item's tag is inline then the compute item is of the form:
185 : *
186 : * \snippet Test_DataBox.cpp compute_item_tag_function
187 : *
188 : * Compute tags can also have their functions be overloaded on the type of its
189 : * arguments:
190 : *
191 : * \snippet Test_DataBox.cpp overload_compute_tag_type
192 : *
193 : * or be overloaded on the number of arguments:
194 : *
195 : * \snippet Test_DataBox.cpp overload_compute_tag_number_of_args
196 : *
197 : * Compute tag function templates are implemented as follows:
198 : *
199 : * \snippet Test_DataBox.cpp overload_compute_tag_template
200 : *
201 : * Finally, overloading, function templates, and variadic functions can be
202 : * combined to produce extremely generic compute tags. The below compute tag
203 : * takes as template parameters a parameter pack of integers, which is used to
204 : * specify several of the arguments. The function is overloaded for the single
205 : * argument case, and a variadic function template is provided for the multiple
206 : * arguments case. Note that in practice few compute tags will be this complex.
207 : *
208 : * \snippet Test_BaseTags.cpp compute_template_base_tags
209 : *
210 : * #### Subitems and Prefix Tags
211 : *
212 : * A simple or compute tag might also hold a collection of data, such as a
213 : * container of `Tensor`s. In many cases you will want to be able to retrieve
214 : * individual elements of the collection from the DataBox without having to
215 : * first retrieve the collection. The infrastructure that allows for this is
216 : * called *Subitems*. The subitems of the parent tag must refer to a subset of
217 : * the data inside the parent tag, e.g. one `Tensor` in the collection. If the
218 : * parent tag is `Parent` and the subitems tags are `Sub<0>, Sub<1>`, then when
219 : * `Parent` is added to the DataBox, so are `Sub<0>` and `Sub<1>`. This means
220 : * the retrieval mechanisms described below will work on `Parent`, `Sub<0>`, and
221 : * `Sub<1>`.
222 : *
223 : * Subitems specify requirements on the tags they act on. For example, there
224 : * could be a requirement that all tags with a certain type are to be treated as
225 : * a Subitems. Let's say that the `Parent` tag holds a `Variables`, and
226 : * `Variables` can be used with the Subitems infrastructure to add the nested
227 : * `Tensor`s. Then all tags that hold a `Variables` will have their subitems
228 : * added into the DataBox. To add a new type as a subitem the `db::Subitems`
229 : * struct must be specialized. See the documentation of `db::Subitems` for more
230 : * details.
231 : *
232 : * The DataBox also supports *prefix tags*, which are commonly used for items
233 : * that are related to a different item by some operation. Specifically, say
234 : * you have a tag `MyTensor` and you want to also have the time derivative of
235 : * `MyTensor`, then you can use the prefix tag `dt` to get `dt<MyTensor>`. The
236 : * benefit of a prefix tag over, say, a separate tag `dtMyTensor` is that prefix
237 : * tags can be added and removed by the compute tags acting on the original tag.
238 : * Prefix tags can also be composed, so a second time derivative would be
239 : * `dt<dt<MyTensor>>`. The net result of the prefix tags infrastructure is that
240 : * the compute tag that returns `dt<MyTensor>` only needs to know its input
241 : * tags, it knows how to name its output based off that. In addition to the
242 : * normal things a simple or a compute tag must hold, prefix tags must have a
243 : * nested type alias `tag`, which is the tag being prefixed. Prefix tags must
244 : * also inherit from `db::PrefixTag` in addition to inheriting from
245 : * `db::SimpleTag` or `db::ComputeTag`.
246 : *
247 : * #### Creating a DataBox
248 : *
249 : * You should never call the constructor of a DataBox directly. DataBox
250 : * construction is quite complicated and the helper functions `db::create` and
251 : * `db::create_from` should be used instead. `db::create` is used to construct a
252 : * new DataBox. It takes two typelists as explicit template parameters, the
253 : * first being a list of the simple tags to add and the second being a list of
254 : * compute tags to add. If no compute tags are being added then only the simple
255 : * tags list must be specified. The tags lists should be passed as
256 : * `db::create<db::AddSimpleTags<simple_tags...>,
257 : * db::AddComputeTags<compute_tags...>>`. The arguments to `db::create` are the
258 : * initial values of the simple tags and must be passed in the same order as the
259 : * tags in the `db::AddSimpleTags` list. If the type of an argument passed to
260 : * `db::create` does not match the type of the corresponding simple tag a static
261 : * assertion will trigger. Here is an example of how to use `db::create`:
262 : *
263 : * \snippet Test_DataBox.cpp create_databox
264 : *
265 : * To create a new DataBox from an existing one use the `db::create_from`
266 : * function. The only time a new DataBox needs to be created is when tags need
267 : * to be removed or added. Like `db::create`, `db::create_from` also takes
268 : * typelists as explicit template parameter. The first template parameter is the
269 : * list of tags to be removed, which is passed using `db::RemoveTags`, second is
270 : * the list of simple tags to add, and the third is the list of compute tags to
271 : * add. If tags are only removed then only the first template parameter needs to
272 : * be specified. If tags are being removed and only simple tags are being added
273 : * then only the first two template parameters need to be specified. Here is an
274 : * example of removing a tag or compute tag:
275 : *
276 : * \snippet Test_DataBox.cpp create_from_remove
277 : *
278 : * Adding a simple tag is done using:
279 : *
280 : * \snippet Test_DataBox.cpp create_from_add_item
281 : *
282 : * Adding a compute tag is done using:
283 : *
284 : * \snippet Test_DataBox.cpp create_from_add_compute_item
285 : *
286 : * #### Accessing and Mutating Items
287 : *
288 : * To retrieve an item from a DataBox use the `db::get` function. `db::get`
289 : * will always return a `const` reference to the object stored in the DataBox
290 : * and will also have full type information available. This means you are able
291 : * to use `const auto&` when retrieving tags from the DataBox. For example,
292 : * \snippet Test_DataBox.cpp using_db_get
293 : *
294 : * If you want to mutate the value of a simple item in the DataBox use
295 : * `db::mutate`. Any compute item that depends on the mutated item will have its
296 : * cached value invalidated and be recomputed the next time it is retrieved from
297 : * the DataBox. `db::mutate` takes a parameter pack of tags to mutate as
298 : * explicit template parameters, a `gsl::not_null` of the DataBox whose items
299 : * will be mutated, an invokable, and extra arguments to forward to the
300 : * invokable. The invokable takes the arguments passed from the DataBox by
301 : * `const gsl::not_null` while the extra arguments are forwarded to the
302 : * invokable. The invokable is not allowed to retrieve anything from the
303 : * DataBox, so any items must be passed as extra arguments using `db::get` to
304 : * retrieve them. For example,
305 : *
306 : * \snippet Test_DataBox.cpp databox_mutate_example
307 : *
308 : * In addition to retrieving items using `db::get` and mutating them using
309 : * `db::mutate`, there is a facility to invoke an invokable with tags from the
310 : * DataBox. `db::apply` takes a `tmpl::list` of tags as an explicit template
311 : * parameter, will retrieve all the tags from the DataBox passed in and then
312 : * invoke the invokable with the items in the tag list. Similarly,
313 : * `db::mutate_apply` invokes the invokable but allows for mutating some of
314 : * the tags. See the documentation of `db::apply` and `db::mutate_apply` for
315 : * examples of how to use them.
316 : *
317 : * #### The Base Tags Mechanism
318 : *
319 : * Retrieving items by tags should not require knowing whether the item being
320 : * retrieved was computed using a compute tag or simply added using a simple
321 : * tag. The framework that handles this falls under the umbrella term
322 : * *base tags*. The reason is that a compute tag can inherit from a simple tag
323 : * with the same item type, and then calls to `db::get` with the simple tag can
324 : * be used to retrieve the compute item as well. That is, say you have a compute
325 : * tag `ArrayCompute` that derives off of the simple tag `Array`, then you can
326 : * retrieve the compute tag `ArrayCompute` and `Array` by calling
327 : * `db::get<Array>(box)`. The base tags mechanism requires that only one `Array`
328 : * tag be present in the DataBox, otherwise a static assertion is triggered.
329 : *
330 : * The inheritance idea can be generalized further with what are called base
331 : * tags. A base tag is an empty `struct` that inherits from `db::BaseTag`. Any
332 : * simple or compute item that derives off of the base tag can be retrieved
333 : * using `db::get`. Consider the following `VectorBase` and `Vector` tag:
334 : *
335 : * \snippet Test_BaseTags.cpp vector_base_definitions
336 : *
337 : * It is possible to retrieve `Vector<1>` from the DataBox using
338 : * `VectorBase<1>`. Most importantly, base tags can also be used in compute tag
339 : * arguments, as follows:
340 : *
341 : * \snippet Test_BaseTags.cpp compute_template_base_tags
342 : *
343 : * As shown in the code example, the base tag mechanism works with function
344 : * template compute tags, enabling generic programming to be combined with the
345 : * lazy evaluation and automatic dependency analysis offered by the DataBox. To
346 : * really demonstrate the power of base tags, let's also have `ArrayComputeBase`
347 : * inherit from a simple tag `Array`, which inherits from a base tag `ArrayBase`
348 : * as follows:
349 : *
350 : * \snippet Test_BaseTags.cpp array_base_definitions
351 : *
352 : * To start, let's create a DataBox that holds a `Vector<0>` and an
353 : * `ArrayComputeBase<0>` (the concrete tag must be used when creating the
354 : * DataBox, not the base tags), retrieve the tags using the base tag mechanism,
355 : * including mutating `Vector<0>`, and then verifying that the dependencies are
356 : * handled correctly.
357 : *
358 : * \snippet Test_BaseTags.cpp base_simple_and_compute_mutate
359 : *
360 : * Notice that we are able to retrieve `ArrayComputeBase<0>` with `ArrayBase<0>`
361 : * and `Array<0>`. We were also able to mutate `Vector<0>` using
362 : * `VectorBase<0>`.
363 : *
364 : * We can even remove tags using their base tags with `db::create_from`:
365 : *
366 : * \snippet Test_BaseTags.cpp remove_using_base
367 : *
368 : * The base tags infrastructure even works with Subitems. Even if you mutate the
369 : * subitem of a parent using a base tag, the appropriate compute item caches
370 : * will be invalidated.
371 : *
372 : * \note All of the base tags infrastructure works for `db::get`, `db::mutate`,
373 : * `db::apply` and `db::mutate_apply`.
374 : */
375 :
376 : /*!
377 : * \defgroup DataBoxTagsGroup DataBox Tags
378 : * \brief Structures and metafunctions for labeling the contents of DataBoxes
379 : */
380 :
381 : /*!
382 : * \defgroup DataStructuresGroup Data Structures
383 : * \brief Various useful data structures used in SpECTRE
384 : */
385 :
386 : /*!
387 : * \defgroup DgSubcellGroup DG-Subcell
388 : * \brief Functions and classes specific to the discontinuous Galerkin method
389 : * supplemented with a finite volume or finite difference subcell limiter. Can
390 : * also be thought of as a DG-FD hybrid method.
391 : */
392 :
393 : /*!
394 : * \defgroup DiscontinuousGalerkinGroup Discontinuous Galerkin
395 : * \brief Functions and classes specific to the Discontinuous Galerkin
396 : * algorithm.
397 : */
398 :
399 : /*!
400 : * \defgroup EllipticSystemsGroup Elliptic Systems
401 : * \brief All available elliptic systems
402 : */
403 :
404 : /*!
405 : * \defgroup EquationsOfStateGroup Equations of State
406 : * \brief The various available equations of state
407 : */
408 :
409 : /*!
410 : * \defgroup ErrorHandlingGroup Error Handling
411 : * Macros and functions used for handling errors
412 : */
413 :
414 : /*!
415 : * \defgroup EventsAndTriggersGroup Events and Triggers
416 : * \brief Classes and functions related to events and triggers
417 : */
418 :
419 : /*!
420 : * \defgroup EvolutionSystemsGroup Evolution Systems
421 : * \brief All available evolution systems and information on how to implement
422 : * evolution systems
423 : *
424 : * \details Actions and parallel components may require an evolution system to
425 : * expose the following types:
426 : *
427 : * - `volume_dim`: The number of spatial dimensions
428 : * - `variables_tag`: The evolved variables to compute DG volume contributions
429 : * and fluxes for.
430 : * - `compute_time_derivative`: A struct that computes the bulk contribution to
431 : * the DG discretization of the time derivative. Must expose a `tmpl::list` of
432 : * `argument_tags` and a static `apply` function that takes the following
433 : * arguments in this order:
434 : * - First, the types of the tensors in
435 : * `db::add_tag_prefix<Metavariables::temporal_id::step_prefix, variables_tag>`
436 : * (which represent the time derivatives of the variables) as not-null pointers.
437 : * - The types of the `argument_tags` as constant references.
438 : *
439 : * Actions and parallel components may also require the Metavariables to expose
440 : * the following types:
441 : *
442 : * - `system`: See above.
443 : * - `temporal_id`: A DataBox tag that identifies steps in the algorithm.
444 : * Generally use `Tags::TimeStepId`.
445 : */
446 :
447 : /*!
448 : * \defgroup ExecutablesGroup Executables
449 : * \brief A list of executables and how to use them
450 : *
451 : * <table class="doxtable">
452 : * <tr>
453 : * <th>Executable Name </th><th>Description </th>
454 : * </tr>
455 : * <tr>
456 : * <td> \ref ParallelInfoExecutablePage "ParallelInfo" </td>
457 : * <td> Executable for checking number of nodes, cores, etc.</td>
458 : * </tr>
459 : * </table>
460 : */
461 :
462 : /*!
463 : * \defgroup FileSystemGroup File System
464 : * \brief A light-weight file system library.
465 : */
466 :
467 : /*!
468 : * \defgroup GeneralRelativityGroup General Relativity
469 : * \brief Contains functions used in General Relativistic simulations
470 : */
471 :
472 : /*!
473 : * \defgroup HDF5Group HDF5
474 : * \brief Functions and classes for manipulating HDF5 files
475 : */
476 :
477 : /*!
478 : * \defgroup InitializationGroup Initialization
479 : * \brief Actions and metafunctions used for initialization of parallel
480 : * components.
481 : */
482 :
483 : /*!
484 : * \defgroup LimitersGroup Limiters
485 : * \brief Limiters to control shocks and surfaces in the solution.
486 : */
487 :
488 : /*!
489 : * \defgroup LinearSolverGroup Linear Solver
490 : * \brief Algorithms to solve linear systems of equations
491 : *
492 : * \details In a way, the linear solver is for elliptic systems what time
493 : * stepping is for the evolution code. This is because the DG scheme for an
494 : * elliptic system reduces to a linear system of equations of the type
495 : * \f$Ax=b\f$, where \f$A\f$ is a global matrix representing the DG
496 : * discretization of the problem. Since this is one equation for each node in
497 : * the computational domain it becomes unfeasible to numerically invert the
498 : * global matrix \f$A\f$. Instead, we solve the problem iteratively so that we
499 : * never need to construct \f$A\f$ globally but only need \f$Ax\f$ that can be
500 : * evaluated locally by virtue of the DG formulation. This action of the
501 : * operator is what we have to supply in each step of the iterative algorithms
502 : * implemented here. It is where most of the computational cost goes and usually
503 : * involves computing a volume contribution for each element and communicating
504 : * fluxes with neighboring elements. Since the iterative algorithms typically
505 : * scale badly with increasing grid size, a preconditioner \f$P\f$ is needed
506 : * in order to make \f$P^{-1}A\f$ easier to invert.
507 : *
508 : * \note The smallest possible residual magnitude the linear solver can reach is
509 : * the product between the machine epsilon and the condition number of the
510 : * linear operator that is being inverted. Smaller residuals are numerical
511 : * artifacts. Requiring an absolute or relative residual below this limit will
512 : * likely make the linear solver run until it reaches its maximum number of
513 : * iterations.
514 : *
515 : * \note Remember that when the linear operator \f$A\f$ corresponds to a PDE
516 : * discretization, decreasing the linear solver residual below the
517 : * discretization error will not improve the numerical solution any further.
518 : * I.e. the error \f$e_k=x_k-x_\mathrm{analytic}\f$ to an analytic solution
519 : * will be dominated by the linear solver residual at first, but even if the
520 : * discretization \f$Ax_k=b\f$ was exactly solved after some iteration \f$k\f$,
521 : * the discretization residual
522 : * \f$Ae_k=b-Ax_\mathrm{analytic}=r_\mathrm{discretization}\f$ would still
523 : * remain. Therefore, ideally choose the absolute or relative residual criteria
524 : * based on an estimate of the discretization residual.
525 : *
526 : * In the iterative algorithms we usually don't work with the physical field
527 : * \f$x\f$ directly. Instead we need to apply the operator to an internal
528 : * variable defined by the respective algorithm. This variable is exposed as the
529 : * `LinearSolver::Tags::Operand` prefix, and the algorithm expects that the
530 : * computed operator action is written into
531 : * `db::add_tag_prefix<LinearSolver::Tags::OperatorAppliedTo,
532 : * LinearSolver::Tags::Operand<...>>` in each step.
533 : */
534 :
535 : /// \defgroup LoggingGroup Logging
536 : /// \brief Functions for logging progress of running code
537 :
538 : /// \defgroup MathFunctionsGroup Math Functions
539 : /// \brief Useful analytic functions
540 :
541 : /*!
542 : * \defgroup NumericalAlgorithmsGroup Numerical Algorithms
543 : * \brief Generic numerical algorithms
544 : */
545 :
546 : /*!
547 : * \defgroup NumericalFluxesGroup Numerical Fluxes
548 : * \brief The set of available numerical fluxes
549 : */
550 :
551 : /*!
552 : * \defgroup ObserversGroup Observers
553 : * \brief Observing/writing data to disk.
554 : */
555 :
556 : /*!
557 : * \defgroup OptionGroupsGroup Option Groups
558 : * \brief Tags used for grouping input file options.
559 : *
560 : * An \ref OptionTagsGroup "option tag" can be placed in a group with other
561 : * option tags to give the input file more structure. To assign a group to an
562 : * option tag, set its `group` type alias to a struct that provides a help
563 : * string and may override a static `name()` function:
564 : *
565 : * \snippet Test_Options.cpp options_example_group
566 : *
567 : * A number of commonly used groups are listed here.
568 : *
569 : * See also the \ref dev_guide_option_parsing "option parsing guide".
570 : */
571 :
572 : /*!
573 : * \defgroup OptionParsingGroup Option Parsing
574 : * Things related to parsing YAML input files.
575 : */
576 :
577 : /*!
578 : * \defgroup OptionTagsGroup Option Tags
579 : * \brief Tags used for options parsed from the input file.
580 : *
581 : * These can be stored in the GlobalCache or passed to the `initialize`
582 : * function of a parallel component.
583 : */
584 :
585 : /*!
586 : * \defgroup ParallelGroup Parallelization
587 : * \brief Functions, classes and documentation related to parallelization and
588 : * Charm++
589 : *
590 : * See
591 : * \ref dev_guide_parallelization_foundations "Parallelization infrastructure"
592 : * for details.
593 : */
594 :
595 : /*!
596 : * \defgroup PeoGroup Performance, Efficiency, and Optimizations
597 : * \brief Classes and functions useful for performance optimizations.
598 : */
599 :
600 : /*!
601 : * \defgroup PrettyTypeGroup Pretty Type
602 : * \brief Pretty printing of types
603 : */
604 :
605 : /*!
606 : * \defgroup ProtocolsGroup Protocols
607 : * \brief Classes that define metaprogramming interfaces
608 : *
609 : * See the \ref protocols section of the dev guide for details.
610 : */
611 :
612 : /*!
613 : * \defgroup PythonBindingsGroup Python Bindings
614 : * \brief Classes and functions useful when writing python bindings.
615 : *
616 : * See the \ref spectre_writing_python_bindings "Writing Python Bindings"
617 : * section of the dev guide for details on how to write python bindings.
618 : */
619 :
620 : /*!
621 : * \defgroup SpecialRelativityGroup Special Relativity
622 : * \brief Contains functions used in special relativity calculations
623 : */
624 :
625 : /*!
626 : * \defgroup SpectralGroup Spectral
627 : * Things related to spectral transformations.
628 : */
629 :
630 : // Note: this group is ordered by how it appears in the rendered Doxygen pages
631 : // (i.e., "Spin-weighted..."), rather than the group's name (i.e., "Swsh...").
632 : /*!
633 : * \defgroup SwshGroup Spin-weighted spherical harmonics
634 : * Utilities, tags, and metafunctions for using and manipulating spin-weighted
635 : * spherical harmonics
636 : */
637 :
638 : /*!
639 : * \defgroup SurfacesGroup Surfaces
640 : * Things related to surfaces.
641 : */
642 :
643 : /*!
644 : * \defgroup TensorGroup Tensor
645 : * Tensor use documentation.
646 : */
647 :
648 : /*!
649 : * \defgroup TensorExpressionsGroup Tensor Expressions
650 : * Tensor Expressions allow writing expressions of
651 : * tensors in a way similar to what is used with pen and paper.
652 : *
653 : * Tensor expressions are implemented using (smart) expression templates. This
654 : * allows a domain specific language making expressions such as
655 : * \code
656 : * auto T = evaluate<Indices::_a_t, Indices::_b_t>(F(Indices::_b,
657 : * Indices::_a));
658 : * \endcode
659 : * possible.
660 : */
661 :
662 : /*!
663 : * \defgroup TestingFrameworkGroup Testing Framework
664 : * \brief Classes, functions, macros, and instructions for developing tests
665 : *
666 : * \details
667 : *
668 : * SpECTRE uses the testing framework
669 : * [Catch](https://github.com/philsquared/Catch). Catch supports a variety of
670 : * different styles of tests including BDD and fixture tests. The file
671 : * `cmake/SpectreAddCatchTests.cmake` parses the source files and adds the found
672 : * tests to ctest with the correct properties specified by tags and attributes.
673 : *
674 : * ### Usage
675 : *
676 : * To run the tests, type `ctest` in the build directory. You can specify
677 : * a regex to match the test name using `ctest -R Unit.Blah`, or run all
678 : * tests with a certain tag using `ctest -L tag`.
679 : *
680 : * ### Comparing double-precision results
681 : *
682 : * To compare two floating-point numbers that may differ by round-off, use the
683 : * helper object `approx`. This is an instance of Catch's comparison class
684 : * `Approx` in which the relative tolerance for comparisons is set to roughly
685 : * \f$10^{-14}\f$ (i.e. `std::numeric_limits<double>::%epsilon()*100`).
686 : * When possible, we recommend using `approx` for fuzzy comparisons as follows:
687 : * \example
688 : * \snippet Test_TestingFramework.cpp approx_default
689 : *
690 : * For checks that need more control over the precision (e.g. an algorithm in
691 : * which round-off errors accumulate to a higher level), we recommend using
692 : * the `approx` helper with a one-time tolerance adjustment. A comment
693 : * should explain the reason for the adjustment:
694 : * \example
695 : * \snippet Test_TestingFramework.cpp approx_single_custom
696 : *
697 : * For tests in which the same precision adjustment is re-used many times, a new
698 : * helper object can be created from Catch's `Approx` with a custom precision:
699 : * \example
700 : * \snippet Test_TestingFramework.cpp approx_new_custom
701 : *
702 : * Note: We provide the `approx` object because Catch's `Approx` defaults to a
703 : * very loose tolerance (`std::numeric_limits<float>::%epsilon()*100`, or
704 : * roughly \f$10^{-5}\f$ relative error), and so is poorly-suited to checking
705 : * many numerical algorithms that rely on double-precision accuracy. By
706 : * providing a tighter tolerance with `approx`, we avoid having to redefine the
707 : * tolerance in every test.
708 : *
709 : * ### Attributes
710 : *
711 : * Attributes allow you to modify properties of the test. Attributes are
712 : * specified as follows:
713 : * \code
714 : * // [[TimeOut, 10]]
715 : * // [[OutputRegex, The error message expected from the test]]
716 : * SPECTRE_TEST_CASE("Unit.Blah", "[Unit]") {
717 : * \endcode
718 : *
719 : * Available attributes are:
720 : *
721 : * <table class="doxtable">
722 : * <tr>
723 : * <th>Attribute </th><th>Description </th>
724 : * </tr>
725 : * <tr>
726 : * <td>TimeOut </td>
727 : * <td>override the default timeout and set the timeout to N seconds. This
728 : * should be set very sparingly since unit tests are designed to be
729 : * short. If your test is too long you should consider testing smaller
730 : * portions of the code if possible, or writing an integration test instead.
731 : * </td>
732 : * </tr>
733 : * <tr>
734 : * <td>OutputRegex </td>
735 : * <td>
736 : * When testing failure modes the exact error message must be tested, not
737 : * just that the test failed. Since the string passed is a regular
738 : * expression you must escape any regex tokens. For example, to match
739 : * `some (word) and` you must specify the string `some \(word\) and`.
740 : * If your error message contains a newline, you can match it using the
741 : * dot operator `.`, which matches any character.
742 : * </td>
743 : * </tr>
744 : * </table>
745 : *
746 : * \example
747 : * \snippet Test_H5.cpp willfail_example_for_dev_doc
748 : *
749 : * ### Debugging Tests in GDB or LLDB
750 : *
751 : * Several tests fail intentionally at the executable level to test error
752 : * handling like ASSERT statements in the code. CTest is aware of which
753 : * should fail and passes them. If you want to debug an individual test
754 : * in a debugger you need to run a single test
755 : * using the %RunTests executable (in dg-charm-build/bin/RunTests) you
756 : * must specify the name of the test as the first argument. For example, if you
757 : * want to run just the "Unit.Gradient" test you can run
758 : * `./bin/RunTests Unit.Gradient`. If you are using a debugger launch the
759 : * debugger, for example if you're using LLDB then run `lldb ./bin/RunTests`
760 : * and then to run the executable inside the debugger use `run Unit.Gradient`
761 : * inside the debugger.
762 : */
763 :
764 : /*!
765 : * \defgroup TimeGroup Time
766 : * \brief Code related to the representation of time during simulations.
767 : *
768 : * The time covered by a simulation is divided up into a sequence of
769 : * adjacent, non-overlapping (except at endpoints) intervals referred
770 : * to as "slabs". The boundaries between slabs can be placed at
771 : * arbitrary times. Slabs, as represented in the code as the Slab
772 : * class, provide comparison operators comparing slabs agreeing with
773 : * the definition as a sequence of intervals. Slabs that do not
774 : * jointly belong to any such sequence should not be compared.
775 : *
776 : * The specific time is represented by the Time class, which encodes
777 : * the slab containing the time and the fraction of the slab that has
778 : * elapsed as an exact rational. Times are comparable according to
779 : * their natural time ordering, except for times belonging to
780 : * incomparable slabs.
781 : *
782 : * Differences in time within a slab are represented as exact
783 : * fractions of that slab by the TimeDelta class. TimeDeltas are only
784 : * meaningful within a single slab, with the exception that the ratio
785 : * of objects with different slabs may be taken, resulting in an
786 : * inexact floating-point result. Longer intervals of time are
787 : * represented using floating-point values.
788 : */
789 :
790 : /*!
791 : * \defgroup TimeSteppersGroup Time Steppers
792 : * A collection of ODE integrators primarily used for time stepping.
793 : */
794 :
795 : /*!
796 : * \defgroup TypeTraitsGroup Type Traits
797 : * A collection of useful type traits, including C++14 and C++17 additions to
798 : * the standard library.
799 : */
800 :
801 : /*!
802 : * \defgroup UtilitiesGroup Utilities
803 : * \brief A collection of useful classes, functions and metafunctions.
804 : */
805 :
806 : /*!
807 : * \defgroup VariableFixingGroup Variable Fixing
808 : * \brief A collection of different variable fixers ranging in sophistication.
809 : *
810 : * Build-up of numerical error can cause physical quantities to evolve
811 : * toward non-physical values. For example, pressure and density may become
812 : * negative. This will subsequently lead to failures in numerical inversion
813 : * schemes to recover the corresponding convervative values. A rough fix that
814 : * enforces physical quantities stay physical is to simply change them by hand
815 : * when needed. This can be done at various degrees of sophistication, but in
816 : * general the fixed quantities make up a negligible amount of the physics of
817 : * the simulation; a rough fix is vastly preferred to a simulation that fails
818 : * to complete due to nonphysical quantities.
819 : */
|
