Source code for schola.core.utils

Copied!


# Copyright (c) 2024 Advanced Micro Devices, Inc. All Rights Reserved.
"""
Utility functions and classes for Schola.
"""

from functools import cached_property, singledispatchmethod
from typing import List, Optional, Tuple, TypeVar, Dict, Iterable, Union

K = TypeVar("K")
V = TypeVar("V")
T = TypeVar("T")

# A generic recursive dictionary type
NestedDict = Dict[K,Union[V,"NestedDict[V]"]]

[docs] def nested_get(dct: NestedDict[K,V], keys : Iterable[K], default: V) -> V: """ Get a value from a nested dictionary, returning a default value if the key is not found. Parameters ---------- dct : NestedDict[K,V] The dictionary to search. keys : Iterable[K] The keys to search for in the dictionary. default : V The value to return if the key is not found. Returns ------- V The value found in the dictionary, or the default value if the key is not found. """ curr_dct = dct for key in keys: if key in curr_dct: curr_dct = curr_dct[key] else: return default return curr_dct
[docs] class IdManager: """ A class to manage the mapping between nested and flattened ids. Parameters ---------- ids : List[List[int]] A nested list of lists of ids to manage, index in the list is first id, second id is stored in the second list. Attributes ---------- ids : List[List[int]] The nested list of lists of ids to manage. """
[docs] def __init__(self, ids : List[List[int]]): self.ids = ids
[docs] def flatten_id_dict(self, nested_id_dict: Dict[int, Dict[int, T]], default:Optional[T]=None) -> List[T] : """ Flatten a dictionary of nested ids into a list of values. Parameters ---------- nested_id_dict : Dict[int, Dict[int, T]] The dictionary to flatten. default : Optional[T], optional The default value to use if a key is not found, by default None. Returns ------- List[T] A flattened list of the values found in the dictionary. """ output_list = [default for i in range(0, self.num_ids)] for first_id, nested_ids in nested_id_dict.items(): for second_id, value in nested_ids.items(): output_list[self.id_map[first_id][second_id]] = value return output_list
[docs] def nest_id_list(self, id_list: List[T], default:Optional[T]=None) -> Dict[int, Dict[int, T]]: """ Nest a list of values, indexed by flattened id, into a dictionary of nested ids. Parameters ---------- id_list : List[T] The list of values to convert into a nested dictionary. default : Optional[T], optional The default value to use if a key is not found, by default None. Returns ------- Dict[int, Dict[int, T]] A nested dictionary of the values in `id_list` or `default` if values are missing. """ output_dict = { first_id:{second_id: default for second_id in nested_ids} for first_id, nested_ids in enumerate(self.ids) } for flat_id, body in enumerate(id_list): first_id, second_id = self[flat_id] output_dict[first_id][second_id] = body return output_dict
@singledispatchmethod def __getitem__(self, key): """ Convert a key into a nested or flattened id, from a flattened or nested id respectively. Parameters ---------- key : Union[int, Tuple[int,int]] The key to convert. Returns ------- Union[Tuple[int,int], int] The converted key. Raises ------ NotImplementedError If the key is not of type int or Tuple[int,int]. """ raise NotImplementedError("get item not supported for keys that aren't int or Tuple[int,int]") @__getitem__.register def _(self, key: int) -> Tuple[int,int]: return self.id_list[key] @__getitem__.register def _(self, key : tuple) -> int: assert len(key) == 2, "if supplying tuple key must supply a key of length 2" return self.id_map[key[0]][key[1]]
[docs] def get_nested_id(self,flat_id:int) -> Tuple[int,int]: """ Get the nested id from a flattened id. Parameters ---------- flat_id : int The flattened id to convert. Returns ------- Tuple[int,int] The nested id. """ return self[flat_id]
[docs] def get_flattened_id(self,first_id:int,second_id:int) -> int: """ Get the flattened id from a nested id. Parameters ---------- first_id : int The first id. second_id : int The second id. Returns ------- int The flattened id. """ return self[first_id,second_id]
@cached_property def id_list(self) -> List[Tuple[int, int]]: """ List of nested ids, for lookups from flattened id to nested ids. Returns ------- List[Tuple[int, int]] List of nested ids. """ id_list = [] for first_id, nested_ids in enumerate(self.ids): for second_id in nested_ids: id_list.append((first_id, second_id)) return id_list @cached_property def id_map(self) -> List[Dict[int,int]]: """ List of dictionaries mapping nested ids to flattened ids. Returns ------- List[Dict[int,int]] List of dictionaries mapping nested ids to flattened ids. """ id_map = [{} for first_id in self.ids] uid = 0 for first_id, nested_ids in enumerate(self.ids): for second_id in nested_ids: id_map[first_id][second_id] = uid uid += 1 return id_map
[docs] def partial_get(self,first_id:int) -> List[int]: """ Get the second ids for a given first id. Parameters ---------- first_id : int The first id to get the second ids for. Returns ------- List[int] The second ids for the given first id. """ return self.ids[first_id]
@cached_property def num_ids(self) -> int: """ The number of ids managed by the IdManager. Returns ------- int The number of ids. """ return sum(map(len,self.ids))

Related pages

  • Visit the Schola product page for download links and more information.

Looking for more documentation on GPUOpen?

AMD GPUOpen software blogs

Our handy software release blogs will help you make good use of our tools, SDKs, and effects, as well as sharing the latest features with new releases.

GPUOpen Manuals

Don’t miss our manual documentation! And if slide decks are what you’re after, you’ll find 100+ of our finest presentations here.

AMD GPUOpen Performance Guides

The home of great performance and optimization advice for AMD RDNA™ 2 GPUs, AMD Ryzen™ CPUs, and so much more.

Getting started: AMD GPUOpen software

New or fairly new to AMD’s tools, libraries, and effects? This is the best place to get started on GPUOpen!

AMD GPUOpen Getting Started Development and Performance

Looking for tips on getting started with developing and/or optimizing your game, whether on AMD hardware or generally? We’ve got you covered!

AMD GPUOpen Technical blogs

Browse our technical blogs, and find valuable advice on developing with AMD hardware, ray tracing, Vulkan®, DirectX®, Unreal Engine, and lots more.

Find out more about our software!

AMD GPUOpen Effects - AMD FidelityFX technologies

Create wonder. No black boxes. Meet the AMD FidelityFX SDK!

AMD GPUOpen Samples

Browse all our useful samples. Perfect for when you’re needing to get started, want to integrate one of our libraries, and much more.

AMD GPUOpen developer SDKs

Discover what our SDK technologies can offer you. Query hardware or software, manage memory, create rendering applications or machine learning, and much more!

AMD GPUOpen Developer Tools

Analyze, Optimize, Profile, Benchmark. We provide you with the developer tools you need to make sure your game is the best it can be!