Skip to content

perf: dont use a vstack before indexing #25

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

Open
wants to merge 9 commits into
base: main
Choose a base branch
from
Open

perf: dont use a vstack before indexing #25

Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
23b816e
feat: remove `vstack`
ilan-gold Aug 19, 2025
124dfdb
fix: remove shape hotpath
ilan-gold Aug 19, 2025
89c8233
Merge branch 'ig/remove_shape_hotpath' into ig/no_vstack
ilan-gold Aug 19, 2025
171c472
fix: use array
ilan-gold Aug 19, 2025
4a7518e
feat: cupy early
ilan-gold Aug 21, 2025
946fd65
Merge branch 'main' into ig/no_vstack_cupy
ilan-gold Aug 22, 2025
bcb6016
fix: oops, actually yield properly when batch size matches in-memory
ilan-gold Aug 22, 2025
c446e63
fix: properly handle leftovers
ilan-gold Aug 22, 2025
623bbe7
Merge branch 'main' into ig/no_vstack_cupy
ilan-gold Aug 22, 2025
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
145 changes: 96 additions & 49 deletions arrayloaders/io/zarr_loader.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,8 @@


def split_given_size(a: np.ndarray, size: int) -> list[np.ndarray]:
if size == 1:
return a
return np.split(a, np.arange(size, len(a), size))


Expand Down Expand Up @@ -301,10 +303,14 @@ def iter(
)
# In order to handle data returned where (chunk_size * preload_nchunks) mod batch_size != 0
# we must keep track of the leftover data.
in_memory_data = None
chunks: list[InputInMemoryArray] = []
in_memory_labels = None
in_memory_indices = None
mod = self.sp_module if issubclass(self.dataset_type, ad.abc.CSRDataset) else np
vstack = (
self.sp_module.vstack
if issubclass(self.dataset_type, ad.abc.CSRDataset)
else np.vstack
)
for chunk_indices in _batched(
self._get_chunks(chunk_size, worker_handle, shuffle), preload_nchunks
):
Expand All @@ -317,18 +323,9 @@ def iter(
]
dataset_index_to_slices = self._slices_to_slices_with_array_index(slices)
# Fetch the data over slices
chunks: list[InputInMemoryArray] = zsync.sync(
index_datasets(dataset_index_to_slices, fetch_data)
)
if any(isinstance(c, CSRContainer) for c in chunks):
chunks_converted: list[OutputInMemoryArray] = [
self.sp_module.csr_matrix(
tuple(self.np_module.asarray(e) for e in c.elems), shape=c.shape
)
for c in chunks
]
else:
chunks_converted = [self.np_module.asarray(c) for c in chunks]
chunks += zsync.sync(index_datasets(dataset_index_to_slices, fetch_data))
chunks_converted = self._to_output_array(chunks)

# Accumulate labels
labels: None | list[np.ndarray] = None
if self.labels is not None:
Expand Down Expand Up @@ -364,11 +361,6 @@ def iter(
for index in dataset_indices
]
# Do batch returns, handling leftover data as necessary
in_memory_data = (
mod.vstack(chunks_converted)
if in_memory_data is None
else mod.vstack([in_memory_data, *chunks_converted])
)
if self.labels is not None:
in_memory_labels = (
np.concatenate(labels)
Expand All @@ -384,41 +376,96 @@ def iter(
# Create random indices into in_memory_data and then index into it
# If there is "leftover" at the end (see the modulo op),
# save it for the next iteration.
batch_indices = np.arange(in_memory_data.shape[0])
if shuffle:
np.random.default_rng().shuffle(batch_indices)
splits = split_given_size(batch_indices, self._batch_size)
for i, s in enumerate(splits):
if s.shape[0] == self._batch_size:
res = [
in_memory_data[s],
in_memory_labels[s] if self.labels is not None else None,
]
if self._return_index:
res += [in_memory_indices[s]]
yield tuple(res)
if i == (
len(splits) - 1
): # end of iteration, leftover data needs be kept
if (s.shape[0] % self._batch_size) != 0:
in_memory_data = in_memory_data[s]
if in_memory_labels is not None:
in_memory_labels = in_memory_labels[s]
if in_memory_indices is not None:
in_memory_indices = in_memory_indices[s]
else:
in_memory_data = None
in_memory_labels = None
in_memory_indices = None
if in_memory_data is not None: # handle any leftover data
if self._batch_size != (num_obs := sum(c.shape[0] for c in chunks)):
batch_indices = np.arange(num_obs)
if shuffle:
np.random.default_rng().shuffle(batch_indices)
splits = split_given_size(batch_indices, self._batch_size)
for i, s in enumerate(splits):
s, chunks_reindexed = self._reindex_against_integer_indices(
s, chunks_converted
)
if s.shape[0] == self._batch_size:
res = [
vstack(chunks_reindexed)
if len(chunks_reindexed) > 1
else chunks_reindexed[0],
in_memory_labels[s] if self.labels is not None else None,
]
if self._return_index:
res += [in_memory_indices[s]]
yield tuple(res)
if i == (
len(splits) - 1
): # end of iteration, leftover data needs be kept
if (s.shape[0] % self._batch_size) != 0:
chunks = chunks_reindexed
if in_memory_labels is not None:
in_memory_labels = in_memory_labels[s]
if in_memory_indices is not None:
in_memory_indices = in_memory_indices[s]
else:
chunks = []
in_memory_labels = None
in_memory_indices = None
elif len(chunks_converted) > 0: # handle batch size matches in-memory
res = [
vstack(chunks_converted),
in_memory_labels if self.labels is not None else None,
]
if self._return_index:
res += [in_memory_indices]
yield tuple(res)
chunks = []
in_memory_labels = None
in_memory_indices = None
if len(chunks) > 0: # handle leftover data
res = [
in_memory_data,
in_memory_labels if self.labels is not None else None,
vstack(self._to_output_array(chunks)),
np.asarray(in_memory_labels) if self.labels is not None else None,
]
if self._return_index:
res += [in_memory_indices]
res += [np.asarray(in_memory_indices)]
yield tuple(res)

def _to_output_array(
self, chunks: list[InputInMemoryArray | OutputInMemoryArray]
) -> list[OutputInMemoryArray]:
if any(isinstance(c, CSRContainer) for c in chunks):
return [
self.sp_module.csr_matrix(
tuple(self.np_module.asarray(e) for e in c.elems), shape=c.shape
)
if isinstance(c, CSRContainer)
else c
for c in chunks
]
elif any(isinstance(c, np.ndarray) for c in chunks):
return [self.np_module.asarray(c) for c in chunks]
return chunks

def _reindex_against_integer_indices(
self, indices: np.ndarray, chunks: list[OutputInMemoryArray]
) -> tuple[np.ndarray, list[OutputInMemoryArray]]:
upper_bounds = np.cumsum(np.array([c.shape[0] for c in chunks]))
lower_bounds = np.concatenate([np.array([0]), upper_bounds[:-1]])
reindexed, chunks_reindexed = list(
zip(
*(
(reindexed, c[self.np_module.asarray(reindexed - lower)])
for c, upper, lower in zip(
chunks, upper_bounds, lower_bounds, strict=False
)
if (
reindexed := indices[(indices < upper) & (indices >= lower)]
).shape[0]
> 0
),
strict=False,
)
)
return np.concatenate(reindexed), list(chunks_reindexed)


AnnDataManager.add_datasets.__doc__ = add_dataset_docstring
AnnDataManager.add_dataset.__doc__ = add_dataset_docstring
Expand Down