Add example of running with example results

README.md

@@ -13,6 +13,10 @@ of options for reading and writing 3D imaging data.
 The final write-up can be found at
 [https://heftieproject.github.io/zarr-benchmarks/](https://heftieproject.github.io/zarr-benchmarks/).
 
+This is based on the results in the `/example_results` directory. To re-create
+all plots locally (inside `data/plots`), follow the installation instructions
+below then run: `python src/zarr_benchmarks/create_plots.py --example_results`.
+
 ## Other related work
 
 - [`zarr-developers/zarr-benchmark`](https://github.com/zarr-developers/zarr-benchmark)

docs/index.md

@@ -30,7 +30,7 @@ These benchmarks are part of the wider
   increase compression ratio:
   - image data, setting it to `"shuffle"`
   - sparse labels, setting it to `"bitshuffle"`
-  - dense labels not setting `shuffle` at all
+  - dense labels, not setting `shuffle` at all
 
 ## Configuration
 
@@ -142,9 +142,9 @@ tensorstore library).
 ![Shuffle vs write time with a shorter write time for shuffle than for no shuffle](assets/shuffle_write.png)
 
 Setting the _shuffle_ configuration to "shuffle" increases the compression ratio
-for imagaing data from ~1.5 to ~1.9, and does not substatially change the read
-or write times. We found that different shuffle options have different outcomes
-for different types of data however.
+for imaging data from ~1.5 to ~1.9, and does not substatially change the read or
+write times. We found that different shuffle options have different outcomes for
+different types of data however.
 
 ### Chunk size