feat(add_field) check type and convert to np.float64 is array type is int

Author: casenave

CHANGELOG.md

@@ -13,6 +13,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Changed
 
+- (containers/features) `add_field` now warns and auto-converts `np.int32`/`np.int64` fields to `np.float64` for CGNS compatibility.
 - (sample) add_tree: add in_place arg to decide whether the added tree should be kept unchanged (with in_place=False).
 - (docs/actions) better explain which package and dev env for which OS, and repair readthedocs generation.
 - (containers/managers) return a warning when an automatically resolved time stamp is different from 0.0.

src/plaid/containers/features.py

@@ -1160,7 +1160,9 @@ def add_field(
 
         Args:
             name (str): The name of the field to be added.
-            field (Field): The field data to be added.
+            field (Field): The field data to be added. Integer arrays with dtype
+                ``np.int32`` or ``np.int64`` are automatically converted to
+                ``np.float64`` (with a warning) for CGNS compatibility.
             location (str, optional): The grid location where the field will be stored. Defaults to 'Vertex'.
                 Possible values : :py:const:`plaid.constants.CGNS_FIELD_LOCATIONS`
             zone_name (str, optional): The name of the zone where the field will be added. Defaults to None.