Buckets:
| import numpy | |
| import pytest | |
| import sparse | |
| from tiled.adapters.mapping import MapAdapter | |
| from tiled.adapters.sparse import COOAdapter | |
| from tiled.client import Context, from_context | |
| from tiled.server.app import build_app | |
| N, M = 3, 5 | |
| state = numpy.random.RandomState(0) | |
| a = state.random((N * 2, M * 2)) | |
| a[a < 0.5] = 0 # Fill half of the array with zeros. | |
| s = sparse.COO(a) | |
| blocks = {} | |
| for i in range(2): | |
| for j in range(2): | |
| chunk = s[i * N : (1 + i) * N, j * M : (1 + j) * M] # noqa: E203 | |
| # Below we test the COOAdapter.from_global_ref constructor, so | |
| # coordinates should be in the reference frame of the whole array. | |
| coords = chunk.coords + [[i * N], [j * M]] | |
| blocks[i, j] = coords, chunk.data | |
| chunks = ((N, N), (M, M)) | |
| dims = ["x", "y"] | |
| mapping = { | |
| "single_chunk": COOAdapter.from_coo(sparse.COO(a), dims=dims), | |
| "multi_chunk": COOAdapter.from_global_ref( | |
| blocks=blocks, shape=(2 * N, 2 * M), chunks=chunks, dims=dims | |
| ), | |
| } | |
| tree = MapAdapter(mapping) | |
| def client(): | |
| app = build_app(tree) | |
| with Context.from_app(app) as context: | |
| client = from_context(context) | |
| yield client | |
| def test_sparse_single_chunk(client): | |
| sc = client["single_chunk"] | |
| actual_via_slice = sc[:] | |
| actual_via_read = sc.read() | |
| actual_via_todense = sc.todense() | |
| actual_via_read_block = sc.read_block((0, 0)) | |
| assert numpy.array_equal(actual_via_slice.todense(), actual_via_read.todense()) | |
| assert numpy.array_equal( | |
| actual_via_slice.todense(), actual_via_read_block.todense() | |
| ) | |
| assert numpy.array_equal(actual_via_slice.todense(), a) | |
| assert numpy.array_equal(actual_via_todense, a) | |
| assert sc.shape == a.shape | |
| assert sc.ndim == a.ndim | |
| assert sc.chunks == tuple((i,) for i in a.shape) | |
| assert sc.dims == dims | |
| def test_sparse_multi_chunk(client): | |
| sc = client["multi_chunk"] | |
| actual_via_slice = sc[:] | |
| actual_via_read = sc.read() | |
| actual_via_todense = sc.todense() | |
| assert numpy.array_equal(actual_via_slice.todense(), actual_via_read.todense()) | |
| assert numpy.array_equal(actual_via_slice.todense(), a) | |
| assert numpy.array_equal(actual_via_todense, a) | |
| assert sc.shape == a.shape | |
| assert sc.ndim == a.ndim | |
| assert sc.chunks == chunks | |
| assert sc.dims == dims | |
Xet Storage Details
- Size:
- 2.34 kB
- Xet hash:
- bffa4e3a5e0eabcb5c195d8734cb966e524d8d17a921426133d96b8a3356065b
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.