Explain key matching in docstring of DensityCorrelations

python/featomic/featomic/clebsch_gordan/_density_correlations.py

@@ -85,7 +85,19 @@ def compute(
 
         where \\rho^{\\nu=1} is the input ``density`` of correlation order 1 (body order
         2), and \\rho^{\\nu=n_{corr} + 1} is the output density of correlation order
-        ``n_correlations + 1``
+        ``n_correlations + 1``. 
+
+        The input ``density`` must have at least ``"o3_lambda"`` and  ``"o3_sigma"``
+        key dimensions, as well as a single components axis with the ``"o3_mu"``
+        dimension. Any other key dimensions present will be matched, while all properties
+        dimensions will have their full tensor product computed.
+
+        For instance, if the key dimensions of ``density`` are ``"o3_lambda"``,
+        ``"o3_sigma"``, ``"center_type"``, and ``"neighbor_type"``, pairs of blocks will
+        only be combined if they have both the same center and neighbor atom type. To
+        fully correlate all combinations of ``"neighbor_type"``, for example, this key
+        dimension should be moved to properties first, using the :py:class:`TensorMap`
+        ``keys_to_properties`` method.
 
         Before performing any correlations, the properties dimensions of ``density`` are
         modified to carry a "_1" suffix. At each iteration, the dimension names of the