write and read hdf5

content/40.units_and_quantities.md

@@ -595,7 +595,26 @@ and `unyt.loadtxt` to read them back in:
 >>> vel, mass = u.loadtxt("sphere.dat", usecols=(1,2), delimiter="\t")
 ```
 
+`unyt_array`s can be written to HDF5 format using the `write_hdf5` method, where 
+one can optionally specify the HDF5 dataset and group where it can be stored:
 
+```python
+>>> # Store the "b" array to the dataset "array_data" at the top of the file
+>>> b.write_hdf5("velocity1.hdf5")
+>>> # Store the "b" array to the dataset "velocity" at the top of the file
+>>> b.write_hdf5("velocity2.hdf5", dataset_name="velocity")
+>>> # Store the "b" array to the dataset "velocity" in the group "fields"
+>>> b.write_hdf5("velocity3.hdf5", dataset_name="velocity", group_name="fields")
+```
+
+and read back in using `from_hdf5`:
+
+```python
+>>> import unyt as u 
+>>> v1 = u.read_hdf5("velocity1.hdf5")
+>>> v2 = u.read_hdf5("velocity2.hdf5", dataset_name="velocity")
+>>> v3 = u.read_hdf5("velocity3.hdf5", dataset_name="velocity", group_name="fields")
+```
 
 #### Exporting `unyt_array`s to other unit packages