docs: add per-data-type tutorials

.gitignore

@@ -26,3 +26,13 @@ __pycache__/
 
 # test-time
 pre-deps.txt
+
+# test datasets
+*.zarr
+*.h5ad
+*.vcf.gz
+*.log
+*.pvar
+*.pgen
+*.psam
+*.tbi

README.md

@@ -165,7 +165,7 @@ If you use `annbatch` in your work, please cite the `annbatch` publication as fo
 
 [scverse]: https://scverse.org/
 
-[in-depth section of our docs]: https://annbatch.readthedocs.io/en/stable/notebooks/example.html
+[in-depth section of our docs]: https://annbatch.readthedocs.io/en/stable/tutorials/quickstart.html
 
 [installation]: https://annbatch.readthedocs.io/en/stable/installation.html
 

docs/detailed-walkthrough.md

@@ -1,7 +1,7 @@
 # Detailed Walkthrough
 
 This page walks through how `annbatch` works in depth.
-For a hands-on, runnable version, see the {doc}`quickstart notebook <notebooks/example>`.
+For a hands-on, runnable version, see the {doc}`quickstart notebook <tutorials/quickstart>`.
 
 ## Preprocessing of On-Disk Data
 

docs/index.md

@@ -10,6 +10,7 @@ into memory, while keeping your GPU fed with high-throughput, shuffled mini-batc
 ```{note}
 You can also use the {class}`annbatch.Loader` on raw {class}`zarr.Array` objects via {meth}`~annbatch.Loader.add_datasets` if your object does not fit the {class}`anndata.AnnData` class object cleanly, as long as the data is semantically row-wise oriented on-disk.
 The {meth}`annbatch.Loader.__iter__` fetching of (contiguous) data is simply done along the first (0th) axis of your data i.e., on-disk zarr with potentially more than two dimensions.
+See the {doc}`single-cell microscopy images tutorial <tutorials/images>` for an example that streams a 4-dimensional image stack this way.
 {class}`~anndata.AnnData` simply provides a convenient wrapper for providing *annotated data* with two dimensions.
 
 Note that the preshuffler {class}`~annbatch.DatasetCollection` requires `AnnData` inputs, and that preshuffling is **highly** recommended for top performance.
@@ -33,12 +34,19 @@ New to *annbatch*? Check out the installation guide and pick the right extras.
 :::
 
 :::{grid-item-card} {octicon}`rocket;1.5em;sd-mr-1` Quickstart
-:link: notebooks/example
+:link: tutorials/quickstart
 :link-type: doc
 
 A hands-on notebook: convert your `.h5ad` files and stream shuffled mini-batches.
 :::
 
+:::{grid-item-card} {octicon}`list-unordered;1.5em;sd-mr-1` Tutorials
+:link: tutorials/scrnaseq
+:link-type: doc
+
+End-to-end runnable tutorials: scRNA-seq, genetics (VCF), microscopy images, and multi-GPU training.
+:::
+
 :::{grid-item-card} {octicon}`book;1.5em;sd-mr-1` User guide
 :link: detailed-walkthrough
 :link-type: doc
@@ -92,7 +100,11 @@ references
 :hidden:
 :maxdepth: 1
 
-notebooks/example
+tutorials/quickstart
+tutorials/scrnaseq
+tutorials/genetics
+tutorials/images
+tutorials/distributed
 ```
 
 ```{toctree}

docs/installation.md

@@ -43,7 +43,7 @@ pip install "annbatch[zarrs,torch,cupy-cuda13]"
 (Replace `cupy-cuda13` with the extra matching your local CUDA version.)
 
 :::{important}
-Always quote the package specifier (`"annbatch[zarrs,torch]"`) and do **not** put spaces between
-the extras. Most shells (bash, zsh) treat the square brackets as glob patterns, so an unquoted
+Always quote the package specifier (`"annbatch[zarrs,torch]"`) and do **not** put spaces between the extras.
+Most shells (bash, zsh) treat the square brackets as glob patterns, so an unquoted
 `annbatch[zarrs,torch]` — or one written as `annbatch[zarrs, torch]` — will fail to install.
 :::