Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat: Add ODD specific metadata #907

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

feat: Add ODD specific metadata #907

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
31 changes: 18 additions & 13 deletions core/include/detray/builders/material_map_builder.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -248,21 +248,26 @@ struct add_sf_material_map {
// The detector only knows the non-owning grid types
using non_owning_t =
typename decltype(mat_grid)::template type<false>;
static_assert(materials_t::template is_defined<non_owning_t>());

// Add the material slabs to the grid
for (const auto& bin : bin_data) {
mat_grid.template populate<replace<>>(bin.local_bin_idx,
bin.single_element);
// Not every mask shape might be used for material maps
if constexpr (materials_t::template is_defined<non_owning_t>()) {
// Add the material slabs to the grid
for (const auto& bin : bin_data) {
mat_grid.template populate<replace<>>(bin.local_bin_idx,
bin.single_element);
}

// Add the material grid to the detector
constexpr auto gid{
materials_t::template get_id<non_owning_t>()};
mat_store.template push_back<gid>(mat_grid);

// Return the index of the new material map
return {gid, static_cast<dindex>(
mat_store.template size<gid>() - 1u)};
} else {
return {materials_t::id::e_none, dindex_invalid};
}

// Add the material grid to the detector
constexpr auto gid{materials_t::template get_id<non_owning_t>()};
mat_store.template push_back<gid>(mat_grid);

// Return the index of the new material map
return {gid,
static_cast<dindex>(mat_store.template size<gid>() - 1u)};
} else {
return {materials_t::id::e_none, dindex_invalid};
}
Expand Down
238 changes: 130 additions & 108 deletions detectors/include/detray/detectors/default_metadata.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,6 +37,15 @@ struct default_metadata {
/// Mask-to-(next)-volume link (potentially switchable for SoA)
using nav_link = std::uint_least16_t;

/// How to store coordinate transform matrices
template <template <typename...> class vector_t = dvector>
using transform_store =
single_store<dtransform3D<algebra_type>, vector_t, geometry_context>;

//
// Surface Primitives
//

/// Mask types
/// @TODO: Need to duplicate for pixel/strip measurement dimensions?
using rectangle = mask<rectangle2D, algebra_type, nav_link>;
Expand Down Expand Up @@ -67,9 +76,52 @@ struct default_metadata {
mask<unbounded<line_square>, algebra_type, nav_link>;
using unmasked_plane = mask<unmasked<>, algebra_type, nav_link>;

/// Material types
/// Give your mask types a name (needs to be consecutive and has to match
/// the types position in the mask store!)
enum class mask_ids : std::uint_least8_t {
e_rectangle2 = 0,
e_portal_rectangle2 = 0,
e_trapezoid2 = 1,
e_annulus2 = 2,
e_cylinder2 = 3,
e_portal_cylinder2 = 4,
e_ring2 = 5,
e_portal_ring2 = 5,
e_straw_tube = 6,
e_drift_cell = 7,
/*e_single1 = 8,
e_single2 = 9,
e_single3 = 10,
e_unbounded_rectangle2 = 11,
e_unbounded_trapezoid2 = 12,
e_unbounded_annulus2 = 13,
e_unbounded_cylinder2 = 14,
e_unbounded_disc2 = 15,
e_unbounded_line_circular2 = 16,
e_unbounded_cell2 = 17,
e_unmasked2 = 18,*/
};

/// How to store masks
template <template <typename...> class tuple_t = dtuple,
template <typename...> class vector_t = dvector>
using mask_store = regular_multi_store<mask_ids, empty_context, tuple_t,
vector_t, rectangle, trapezoid,
annulus, cylinder, cylinder_portal,
disc, straw_tube, drift_cell /*,
single_1, single_2, single_3, unbounded_rectangle, unbounded_trapezoid,
unbounded_annulus, unbounded_cylinder, unbounded_disc, unbounded_straw_tube,
unbounded_cell, unmasked_plane*/>;

//
// Material Description
//

/// Homogeneous material description
/// @{
using slab = material_slab<scalar_t>;
using rod = material_rod<scalar_t>;
/// @}

/// Material grid types (default: closed bounds, regular binning)
/// @{
Expand Down Expand Up @@ -102,6 +154,47 @@ struct default_metadata {

/// @}

/// Give your material types a name (needs to be consecutive and has to
/// match the types position in the mask store!)
enum class material_ids : std::uint_least8_t {
// Material texture (grid) shapes
e_concentric_cylinder2_map = 0u,
e_disc2_map = 1u,
e_cylinder2_map = 2u,
e_rectangle2_map = 3u,
e_trapezoid2_map = 3u,
e_annulus2_map = 0u,
e_drift_cell_map = 5u,
e_straw_tube_map = 5u,
// Volume material
e_cuboid3_map = 7,
e_cylinder3_map = 8u,
// Homogeneous mapetrial
e_slab = 4u,
e_rod = 5u,
e_raw_material = 6u,
e_none = 9u,
};

/// How to store materials
template <template <typename...> class tuple_t = dtuple,
typename container_t = host_container_types>
using material_store = multi_store<
material_ids, empty_context, tuple_t,
grid_collection<concentric_cylinder2_map_t<container_t>>,
grid_collection<disc_map_t<container_t>>,
grid_collection<cylinder2_map_t<container_t>>,
grid_collection<rectangular_map_t<container_t>>,
typename container_t::template vector_type<slab>,
typename container_t::template vector_type<rod>,
typename container_t::template vector_type<material<scalar_t>>,
grid_collection<cuboid_map_t<container_t>>,
grid_collection<cylinder3_map_t<container_t>>>;

//
// Acceleration structures
//

/// surface grid types (default boundaries: closed binning)
/// @TODO: Will we need the types for all grid configurations (binnning,
/// bin boundaries, serializers)?
Expand Down Expand Up @@ -151,85 +244,19 @@ struct default_metadata {

/// @}

/// How to store coordinate transform matrices
template <template <typename...> class vector_t = dvector>
using transform_store =
single_store<dtransform3D<algebra_type>, vector_t, geometry_context>;

/// Give your mask types a name (needs to be consecutive and has to match
/// the types position in the mask store!)
enum class mask_ids : std::uint_least8_t {
e_rectangle2 = 0,
e_portal_rectangle2 = 0,
e_trapezoid2 = 1,
e_annulus2 = 2,
e_cylinder2 = 3,
e_portal_cylinder2 = 4,
e_ring2 = 5,
e_portal_ring2 = 5,
e_straw_tube = 6,
e_drift_cell = 7,
/*e_single1 = 8,
e_single2 = 9,
e_single3 = 10,
e_unbounded_rectangle2 = 11,
e_unbounded_trapezoid2 = 12,
e_unbounded_annulus2 = 13,
e_unbounded_cylinder2 = 14,
e_unbounded_disc2 = 15,
e_unbounded_line_circular2 = 16,
e_unbounded_cell2 = 17,
e_unmasked2 = 18,*/
};

/// How to store masks
template <template <typename...> class tuple_t = dtuple,
template <typename...> class vector_t = dvector>
using mask_store = regular_multi_store<mask_ids, empty_context, tuple_t,
vector_t, rectangle, trapezoid,
annulus, cylinder, cylinder_portal,
disc, straw_tube, drift_cell /*,
single_1, single_2, single_3, unbounded_rectangle, unbounded_trapezoid,
unbounded_annulus, unbounded_cylinder, unbounded_disc, unbounded_straw_tube,
unbounded_cell, unmasked_plane*/>;

/// Give your material types a name (needs to be consecutive and has to
/// match the types position in the mask store!)
enum class material_ids : std::uint_least8_t {
// Material texture (grid) shapes
e_disc2_map = 0u,
e_concentric_cylinder2_map = 1u,
e_cylinder2_map = 2u,
e_rectangle2_map = 3u,
e_trapezoid2_map = 3u,
e_annulus2_map = 0u,
e_drift_cell_map = 5u,
e_straw_tube_map = 5u,
// Volume material
e_cuboid3_map = 7,
e_cylinder3_map = 8u,
// Homogeneous mapetrial
e_slab = 4u,
e_rod = 5u,
e_raw_material = 6u,
e_none = 9u,
/// Acceleration data structures
enum class accel_ids : std::uint_least8_t {
e_brute_force = 0, // test all surfaces in a volume (brute force)
e_cylinder2_grid = 1, // e.g. barrel layers
e_disc_grid = 2, // e.g. endcap layers
e_irr_cylinder2_grid = 3,
e_irr_disc_grid = 4,
// e_cylinder3_grid = 5,
// e_irr_cylinder3_grid = 6,
// ... e.g. frustum navigation types
e_default = e_brute_force,
};

/// How to store materials
template <template <typename...> class tuple_t = dtuple,
typename container_t = host_container_types>
using material_store = multi_store<
material_ids, empty_context, tuple_t,
grid_collection<disc_map_t<container_t>>,
grid_collection<concentric_cylinder2_map_t<container_t>>,
grid_collection<cylinder2_map_t<container_t>>,
grid_collection<rectangular_map_t<container_t>>,
typename container_t::template vector_type<slab>,
typename container_t::template vector_type<rod>,
typename container_t::template vector_type<material<scalar_t>>,
grid_collection<cuboid_map_t<container_t>>,
grid_collection<cylinder3_map_t<container_t>>>;

/// How to link to the entries in the data stores
using transform_link = typename transform_store<>::link_type;
using mask_link = typename mask_store<>::single_link;
Expand All @@ -238,6 +265,28 @@ unbounded_cell, unmasked_plane*/>;
using surface_type =
surface_descriptor<mask_link, material_link, transform_link, nav_link>;

/// How to store the acceleration data structures
template <template <typename...> class tuple_t = dtuple,
typename container_t = host_container_types>
using accelerator_store =
multi_store<accel_ids, empty_context, tuple_t,
brute_force_collection<surface_type, container_t>,
grid_collection<
cylinder2D_sf_grid<surface_type, container_t>>,
grid_collection<disc_sf_grid<surface_type, container_t>>,
grid_collection<
irr_cylinder2D_sf_grid<surface_type, container_t>>,
grid_collection<
irr_disc_sf_grid<surface_type, container_t>> /*,
grid_collection<cylinder3D_sf_grid<surface_type,
container_t>>,
grid_collection<irr_cylinder3D_sf_grid<surface_type,
container_t>>*/>;

//
// Volume descriptors
//

/// How to index the constituent objects (surfaces) in a volume
/// If they share the same index value here, they will be added into the
/// same acceleration data structure (brute force is always at 0)
Expand All @@ -249,41 +298,14 @@ unbounded_cell, unmasked_plane*/>;
e_all = e_size, // i.e. the brute force method and one grid type
};

/// Acceleration data structures
enum class accel_ids : std::uint_least8_t {
e_brute_force = 0, // test all surfaces in a volume (brute force)
e_disc_grid = 1, // e.g. endcap layers
e_cylinder2_grid = 2, // e.g. barrel layers
e_irr_disc_grid = 3,
e_irr_cylinder2_grid = 4,
// e_cylinder3_grid = 5,
// e_irr_cylinder3_grid = 6,
// ... e.g. frustum navigation types
e_default = e_brute_force,
};

/// How a volume links to the accelration data structures
/// In this case: One link for portals/passives and one sensitive surfaces
using object_link_type =
dmulti_index<dtyped_index<accel_ids, dindex>, geo_objects::e_size>;

/// How to store the acceleration data structures
template <template <typename...> class tuple_t = dtuple,
typename container_t = host_container_types>
using accelerator_store =
multi_store<accel_ids, empty_context, tuple_t,
brute_force_collection<surface_type, container_t>,
grid_collection<disc_sf_grid<surface_type, container_t>>,
grid_collection<
cylinder2D_sf_grid<surface_type, container_t>>,
grid_collection<
irr_disc_sf_grid<surface_type, container_t>>,
grid_collection<irr_cylinder2D_sf_grid<
surface_type, container_t>> /*,
grid_collection<cylinder3D_sf_grid<surface_type,
container_t>>,
grid_collection<irr_cylinder3D_sf_grid<surface_type,
container_t>>*/>;
//
// Volume acceleration structure
//

/// Volume search grid
template <typename container_t = host_container_types>
Expand Down
Loading
Loading