- Bioconductor 3.21, May 2025
- Bioconductor 3.21 devel
- Bioconductor 3.20, October 2024
- Fix correctly assign levels to factors in R reader with anndata v0.7 files (Fixes #122)
- Add environment for anndata v0.10.9
- Avoid deprecation warning due to setting
dtype
when creating PythonAnnData
objects - Standardise code styling using {styler}
- Correctly assign levels to factors in R reader (Fixes #122)
- Correctly set
filepath
in the R reader with readingadata.raw
withuse_hdf5 = TRUE
(PR #124 @GabrielHoffman, Fixes #123)
- Skip anndata v0.7.6 tests on Apple Silicon as Python 3.7 is not available
- Fix URL for GTEx 8 tissues dataset in long tests
- Add test using the more complete example H5AD file
- Bioconductor 3.20 devel
- Bioconductor 3.19, May 2023
- Skip anndata v0.7.6 tests on Apple Silicon as Python 3.7 is not available
- Fix URL for GTEx 8 tissues dataset in long tests
- Add test using the more complete example H5AD file
- Prepare
NEWS
for release
- Correctly handle
use_backed = TRUE
with newer anndata versions (Fixes #114) - Improve warnings when converting matrices fails
- Add environment for anndata v0.10.6
- Minor change to writing
DelayedArray
s for compatibility with {HDF5Array} v1.31.1
- Fix the anndata v0.10.2 environment instantiation (Fixes #103)
- Fix a typo in the AnnData Conversion docs (Fixes #100)
- Bioconductor 3.19 devel
- Bioconductor 3.18, October 2023
- Fix the anndata v0.10.2 environment instantiation (Fixes #103)
- Fix a typo in the AnnData Conversion docs (Fixes #100)
- Add environment for anndata v0.10.2
- Add environment for anndata v0.9.2
- Changes for compatibility with {rhdf5} v2.45.1
- Support for enum types that simplifies reading of nullable types in the native R reader
- Pass correct dimensions when converting
raw
(Fixes #96) - Convert anndata backed sparse matrices in
AnnData2SCE()
(Fixes #96)
- Bioconductor 3.18 devel
- Bioconductor 3.17, April 2023
- Pass correct dimensions when converting
raw
(Fixes #96) - Convert anndata backed sparse matrices in
AnnData2SCE()
(Fixes #96)
- Add functions for converting pandas arrays used by anndata when arrays have missing values (Fixes #87)
- Read the correct index names in the R reader (PR #93 mtmorgan)
- Adjust tests to match reader changes
- Add @rcannood as a contributor (PR #90 @rcannood, fixes #88)
- Add compatibility with the anndata v0.8 H5AD format to the the native R writer (PR #86 @jackkamm, fixes #78)
- Bioconductor 3.17 devel
- Bioconductor 3.16, November 2022
- Improve compatibility with the R {anndata} package (PR #76 @rcannood,
fixes #75)
- Python objects are now explicitly converted rather than relying on automatic conversion
- Other minor modifications for compatibility
- Added support for numpy recarrays (dtype number 20) (PR #81, fixes #45,
#28)
- Added a new
py_to_r.numpy.ndarray()
function which extends the default {reticulate} function
- Added a new
- Improvements to warnings
- Improvements and updates to tests
- Pin python version to 3.7.10 in anndata v0.7.6 environment (3.7.12 was not compatible with other dependencies)
- Pin python version to 3.7.12 in anndata v0.7.6 environment to match {basilisk} changes
- Minor changes for compatibility with {cli} v3.4.0
- Added tests for
verbose=TRUE
- Added tests for
- Minor changes for compatibility with the upcoming {Matrix} 1.4-2 release
- Move verbose from
zellkonverterAnnDataEnv()
(Fixes #66)
- Instantiate environments for
basilisk::configureBasiliskEnv()
(Fixes #66) - Allow missing obs/var names when
use_hdf5 = TRUE
(Fixes #65)
- Fix bug in long tests
- Bioconductor 3.16 devel
- Bioconductor 3.15, April 2022
- Minor changes for compatibility with {cli} v3.4.0
- Added tests for
verbose=TRUE
- Added tests for
- Minor changes for compatibility with the upcoming {Matrix} 1.4-2 release
- Move verbose from
zellkonverterAnnDataEnv()
(Fixes #66)
- Instantiate environments for
basilisk::configureBasiliskEnv()
(Fixes #66) - Allow missing obs/var names when
use_hdf5 = TRUE
(Fixes #65)
- Fix bug in long tests
- Fix progress messages in
.convert_anndata_df()
- Allow
data.frames
invarm
inSCE2AnnData()
- Standardise
uns
names to match R conventions inAnnData2SCE()
- Adjust long tests
- Reduce scipy version to 1.7.3
- scipy >= 1.8.0 is incompatible with {reticulate} <= 1.24 (see rstudio/reticulate#1173)
- Add GTEX 8 tissues dataset to long tests (see #58)
- Update the default Python environment to use anndata v0.8.0
- anndata 0.8.0
- h5py 3.6.0
- hdf5 1.12.1
- natsort 8.1.0
- numpy 1.22.3
- packaging 21.3
- pandas 1.4.2
- python 3.8.13
- scipy 1.8.0
- sqlite 3.38.2
- Add options to choose Python environments with different versions of
anndata
- To facilitate this
zellkonverterAnnDataEnv()
andAnnDataDependencies()
are new functions rather than variables - Added a new
.AnnDataVersions
variable which stores the available anndata versions - Updates to the vignette and function documentation explaining this option
- To facilitate this
- Modify how Pandas DataFrames are converted to R
- Columns should now use R approved names with a warning when changes are made
- Bioconductor 3.15 devel
- Bioconductor 3.14, October 2021
- Add progress messages to various functions
- Can be controlled by function arguments or a global variable
- Split
konverter.R
into two files (AnnData2SCE.R
andSCE2AnnData.R
) - Add arguments to control how slots are converted in
AnnData2SCE()
andSCE2AnnData()
(Fixes #47)- Each slot can now be fully converted, skipped entirely or only selected items converted.
- Add support for converting the
raw
slot to analtExp
inAnnData2SCE()
(Fixes #53, fixes #57)
- Add recursive conversion of lists in
AnnData2SCE()
- Correctly handle
DataFrame
objects stored inadata.obsm
- Remove pandas indexes from converted
DataFrame
objects - Add functions for validating
SingleCellExperiment
objects (for testing) - Add long tests for various public datasets
- Fix bug in converting
dgRMatrix
sparse matrices (Fixes #55)
- Bioconductor 3.14 devel
- Bioconductor 3.13, May 2021
- Fix bug in converting
dgRMatrix
sparse matrices (Fixes #55)
- Add experimental native R reader to
readH5AD()
- Update NEWS for release
AnnData2SCE()
no longer returnsdgRMatrix
sparse matrices (Fixes #34)
- Add conversion checks to all slots in
AnnData2SCE()
(See #45) - Enable return conversion of
varm
inSCE2AnnData()
(Fixes #43) - Store
X_name
inAnnData2SCE()
for use bySCE2AnnData()
and add anX_name
argument toAnnData2SCE()
andreadH5AD()
(Fixes #7)
- Add
compression
argument towriteH5AD()
(Fixes #49) - Update anndata Python dependencies, now using anndata v0.7.6
- Adapt to changes in
HDF5Array::HDF5Array()
- Better support for anndata
SparseDataset
arrays (PR #41, Fixes #37, Fixes #42) - More consistent conversion of
metadata
touns
inSCE2AnnData()
(Fixes #40) - Add handling of list columns in
colData
androwData
inSCE2AnnData()
(Fixes #26) - Export
zellkonverterAnnDataEnv
(Fixes #38)
- Handle writing DelayedArray assays on the R side in
writeH5AD()
(PR #35, Fixes #32)
- Adjust
SCE2AnnData()
example (Fixes #31)
- Improved support for HDF5 backed conversion (PR #29, fixes #13)
- Add
example_anndata.h5ad
file toinst/extdata/
and creation script toinst/scripts/
- Improve conversion checks when converting
.uns
tometadata
- Avoid converting
obsp
andvarp
to dense matrices
- Bioconductor 3.13 devel
- Bioconductor 3.12, October 2020
- Avoid converting
obsp
andvarp
to dense matrices
- Merge remaining commits for HDF5 conversion (fixes #33)
- Improved support for HDF5 backed conversion (PR #29, fixes #13, fixes #33)
- Update Python dependencies
- numpy 1.18.5 -> 1.19.1
- pandas 1.0.4 -> 1.1.2
- scipy 1.4.1 -> 1.5.2
- sqlite 3.30.1 -> 3.33.0
- Document character to factor coercion in
writeH5ad()
(Fixes #6) - Add
X_name
argument towriteH5AD()
(Fixes #23)
- Tidy NEWS files for Bioconductor release
- Bump anndata version to 0.7.4
- Document the
krumsiek11.h5ad
file - Remove the
internal
keyword from thezellkonverter-package
documentation
- Update
.gitignore
- Fix SCE to AnnData map figure in PDF manual
- Use
expect_equal()
instead ofexpect_identical()
inwriteH5AD()
sparse matrices test - Edit package title and description
- Initial Bioconductor submission
- Add biocViews to DESCRIPTION
- Edit package description
- Tidy code
- Replace 1:... with
seq_len()
- Add check for scRNAseq in examples (Fixes #18)
- Skip
AnnData
matrices without a transposable R counterpart - Only replace skipped matrices when
use_hdf5 = TRUE
inreadH5AD()
(Fixes #12) - Additional tests for sparse matrices
- Allow assay skipping when converting from
SingleCellExperiment
toAnnData
- Allow skipping of assays that aren't numpy friendly in
writeH5AD()
- Wait for basilisk process shutdown to release
.h5ad
file - Updates to documentation and tests
- Improve conversion between
SingleCellExperiment
andAnnData
(See #8)- Convert between
metadata
anduns
(where objects are compatible) - Convert between
rowPairs
andvarp
- Convert between
colPairs
andobsp
- Convert from
varm
torowData
(but not in reverse)
- Convert between
- Add mapping table to docs
- Tidy documentation and code
- Tidy vignette
- Support for HDF5Array outputs in
readH5AD()
(Fixes #4)
- Avoid checking column names for
colData
androwData
inSCE2AnnData()
- Make sure that all matrices passes to {reticulate} are numpy friendly
- Add more tests
- Update vignette front matter
- Add vignette
- Add examples and improve documentation
- Export
.AnnDataDependencies
for external use
- Add
SCE2AnnData()
function - Add
writeH5AD()
function
- Use internal function in
readH5AD()
- Rename
adata2SCE()
toAnnData2SCE()
- Remove {basilisk} context from
AnnData2SCE()
(See #1)- Now uses the calling context
- Pin more AnnData dependencies (See #1)
- Add test
.h5ad
file - Add test for
readH5AD()
- Add package man page
- Add
adata2SCE()
function - Add
readH5AD()
function
- Add {basilisk} infrastructure
- Set up package