Return the coordinate label of the maximum value along a dimension. Returns a new `DataArray` named after the dimension with the values of the coordinate labels along that dimension corresponding to maximum values along that dimension. In comparison to :py:meth:`~Da
(
self,
dim: Hashable = None,
*,
skipna: bool | None = None,
fill_value: Any = dtypes.NA,
keep_attrs: bool | None = None,
)
| 6104 | ) |
| 6105 | |
| 6106 | def idxmax( |
| 6107 | self, |
| 6108 | dim: Hashable = None, |
| 6109 | *, |
| 6110 | skipna: bool | None = None, |
| 6111 | fill_value: Any = dtypes.NA, |
| 6112 | keep_attrs: bool | None = None, |
| 6113 | ) -> Self: |
| 6114 | """Return the coordinate label of the maximum value along a dimension. |
| 6115 | |
| 6116 | Returns a new `DataArray` named after the dimension with the values of |
| 6117 | the coordinate labels along that dimension corresponding to maximum |
| 6118 | values along that dimension. |
| 6119 | |
| 6120 | In comparison to :py:meth:`~DataArray.argmax`, this returns the |
| 6121 | coordinate label while :py:meth:`~DataArray.argmax` returns the index. |
| 6122 | |
| 6123 | Parameters |
| 6124 | ---------- |
| 6125 | dim : Hashable, optional |
| 6126 | Dimension over which to apply `idxmax`. This is optional for 1D |
| 6127 | arrays, but required for arrays with 2 or more dimensions. |
| 6128 | skipna : bool or None, default: None |
| 6129 | If True, skip missing values (as marked by NaN). By default, only |
| 6130 | skips missing values for ``float``, ``complex``, and ``object`` |
| 6131 | dtypes; other dtypes either do not have a sentinel missing value |
| 6132 | (``int``) or ``skipna=True`` has not been implemented |
| 6133 | (``datetime64`` or ``timedelta64``). |
| 6134 | fill_value : Any, default: NaN |
| 6135 | Value to be filled in case all of the values along a dimension are |
| 6136 | null. By default this is NaN. The fill value and result are |
| 6137 | automatically converted to a compatible dtype if possible. |
| 6138 | Ignored if ``skipna`` is False. |
| 6139 | keep_attrs : bool or None, optional |
| 6140 | If True, the attributes (``attrs``) will be copied from the |
| 6141 | original object to the new one. If False, the new object |
| 6142 | will be returned without attributes. |
| 6143 | |
| 6144 | Returns |
| 6145 | ------- |
| 6146 | reduced : DataArray |
| 6147 | New `DataArray` object with `idxmax` applied to its data and the |
| 6148 | indicated dimension removed. |
| 6149 | |
| 6150 | See Also |
| 6151 | -------- |
| 6152 | Dataset.idxmax, DataArray.idxmin, DataArray.max, DataArray.argmax |
| 6153 | |
| 6154 | Examples |
| 6155 | -------- |
| 6156 | >>> array = xr.DataArray( |
| 6157 | ... [0, 2, 1, 0, -2], dims="x", coords={"x": ["a", "b", "c", "d", "e"]} |
| 6158 | ... ) |
| 6159 | >>> array.max() |
| 6160 | <xarray.DataArray ()> Size: 8B |
| 6161 | array(2) |
| 6162 | >>> array.argmax(...) |
| 6163 | {'x': <xarray.DataArray ()> Size: 8B |