# xarray.core.groupby.DatasetGroupBy.first

#### DatasetGroupBy.first(skipna=None, keep_attrs=None)[[source]](https://github.com/pydata/xarray/blob/main/xarray/core/groupby.py#L1565-L1584)

Return the first element of each group along the group dimension

* **Parameters:**
  * **skipna** ([`bool`](https://docs.python.org/3/builtins/functions.html#bool) or [`None`](https://docs.python.org/3/builtins/constants.html#None), *optional*) – If True, skip missing values (as marked by NaN). By default, only
    skips missing values for float dtypes; other dtypes either do not
    have a sentinel missing value (int) or `skipna=True` has not been
    implemented (object, datetime64 or timedelta64).
  * **keep_attrs** ([`bool`](https://docs.python.org/3/builtins/functions.html#bool) or [`None`](https://docs.python.org/3/builtins/constants.html#None), *optional*) – If True, `attrs` will be copied from the original
    object to the new one.  If False, the new object will be
    returned without attributes.