**Breaking**: data_kind: data is None and required now returns the 'empty' kind

pygmt/clib/session.py

@@ -1767,24 +1767,22 @@ def virtualfile_in(  # noqa: PLR0912
             if check_kind == "raster":
                 valid_kinds += ("grid", "image")
             elif check_kind == "vector":
-                valid_kinds += ("matrix", "vectors", "geojson")
+                valid_kinds += ("empty", "matrix", "vectors", "geojson")
             if kind not in valid_kinds:
                 raise GMTInvalidInput(
                     f"Unrecognized data type for {check_kind}: {type(data)}"
                 )
 
         # Decide which virtualfile_from_ function to use
         _virtualfile_from = {
-            "file": contextlib.nullcontext,
             "arg": contextlib.nullcontext,
+            "empty": self.virtualfile_from_vectors,
+            "file": contextlib.nullcontext,
             "geojson": tempfile_from_geojson,
             "grid": self.virtualfile_from_grid,
             "image": tempfile_from_image,
             "stringio": self.virtualfile_from_stringio,
-            # Note: virtualfile_from_matrix is not used because a matrix can be
-            # converted to vectors instead, and using vectors allows for better
-            # handling of string type inputs (e.g. for datetime data types)
-            "matrix": self.virtualfile_from_vectors,
+            "matrix": self.virtualfile_from_matrix,
             "vectors": self.virtualfile_from_vectors,
         }[kind]
 
@@ -1800,30 +1798,33 @@ def virtualfile_in(  # noqa: PLR0912
                 )
                 warnings.warn(message=msg, category=RuntimeWarning, stacklevel=2)
             _data = (data,) if not isinstance(data, pathlib.PurePath) else (str(data),)
-        elif kind == "vectors":
+        elif kind == "empty":
+            # data is None, so data must be given via x/y/z.
             _data = [x, y]
             if z is not None:
                 _data.append(z)
             if extra_arrays:
                 _data.extend(extra_arrays)
-        elif kind == "matrix":  # turn 2-D arrays into list of vectors
+        elif kind == "vectors":
             if hasattr(data, "items") and not hasattr(data, "to_frame"):
                 # pandas.DataFrame or xarray.Dataset types.
                 # pandas.Series will be handled below like a 1-D numpy.ndarray.
                 _data = [array for _, array in data.items()]
-            elif hasattr(data, "ndim") and data.ndim == 2 and data.dtype.kind in "iuf":
-                # Just use virtualfile_from_matrix for 2-D numpy.ndarray
-                # which are signed integer (i), unsigned integer (u) or
-                # floating point (f) types
-                _virtualfile_from = self.virtualfile_from_matrix
-                _data = (data,)
             else:
                 # Python list, tuple, numpy.ndarray, and pandas.Series types
                 _data = np.atleast_2d(np.asanyarray(data).T)
+        elif kind == "matrix":
+            # GMT can only accept a 2-D matrix which are signed integer (i), unsigned
+            # integer (u) or floating point (f) types. For other data types, we need to
+            # use virtualfile_from_vectors instead, which turns the matrix into list of
+            # vectors and allows for better handling of string type inputs (e.g. for
+            # datetime data types).
+            _data = (data,)
+            if data.dtype.kind not in "iuf":
+                _virtualfile_from = self.virtualfile_from_vectors
 
         # Finally create the virtualfile from the data, to be passed into GMT
         file_context = _virtualfile_from(*_data)
-
         return file_context
 
     def virtualfile_from_data(

pygmt/helpers/utils.py

@@ -14,6 +14,7 @@
 from collections.abc import Iterable, Mapping, Sequence
 from typing import Any, Literal
 
+import numpy as np
 import xarray as xr
 from pygmt.encodings import charset
 from pygmt.exceptions import GMTInvalidInput
@@ -190,7 +191,7 @@ def _check_encoding(
 def data_kind(
     data: Any, required: bool = True
 ) -> Literal[
-    "arg", "file", "geojson", "grid", "image", "matrix", "stringio", "vectors"
+    "arg", "empty", "file", "geojson", "grid", "image", "matrix", "stringio", "vectors"
 ]:
     r"""
     Check the kind of data that is provided to a module.
@@ -200,15 +201,20 @@ def data_kind(
 
     - ``"arg"``: ``data`` is ``None`` and ``required=False``, or bool, int, float,
       representing an optional argument, used for dealing with optional virtual files
+    - ``"empty"`: ``data`` is ``None`` and ``required=True``. It means the data is given
+      via a series of vectors like x/y/z
     - ``"file"``: a string or a :class:`pathlib.PurePath` object or a sequence of them,
       representing one or more file names
     - ``"geojson"``: a geo-like Python object that implements ``__geo_interface__``
       (e.g., geopandas.GeoDataFrame or shapely.geometry)
     - ``"grid"``: a :class:`xarray.DataArray` object that is not 3-D
     - ``"image"``: a 3-D :class:`xarray.DataArray` object
     - ``"stringio"``: a :class:`io.StringIO` object
-    - ``"matrix"``: anything else that is not ``None``
-    - ``"vectors"``: ``data`` is ``None`` and ``required=True``
+    - ``"matrix"``: a 2-D :class:`numpy.ndarray` object
+    - ``"vectors"``: ``data`` is ``None`` and ``required=True``, or any unrecognized
+      data. Common data types include, a :class:`pandas.DataFrame` object, a dictionary
+      with array-like values, a 1-D/3-D :class:`numpy.ndarray` object, or array-like
+      objects.
 
     Parameters
     ----------
@@ -238,6 +244,11 @@ def data_kind(
     >>> data_kind(data=None, required=False)
     'arg'
 
+    The "empty" kind:
+
+    >>> data_kind(data=None)
+    'empty'
+
     The "file" kind:
 
     >>> [data_kind(data=data) for data in ("file.txt", ("file1.txt", "file2.txt"))]
@@ -268,31 +279,31 @@ def data_kind(
 
     The "matrix"`` kind:
 
-    >>> data_kind(data=np.arange(10))  # 1-D numpy.ndarray
-    'matrix'
     >>> data_kind(data=np.arange(10).reshape((5, 2)))  # 2-D numpy.ndarray
     'matrix'
+
+    The "vectors" kind:
+
+    >>> data_kind(data=np.arange(10))  # 1-D numpy.ndarray
+    'vectors'
     >>> data_kind(data=np.arange(60).reshape((3, 4, 5)))  # 3-D numpy.ndarray
-    'matrix'
+    'vectors'
     >>> data_kind(xr.DataArray(np.arange(12), name="x").to_dataset())  # xarray.Dataset
-    'matrix'
+    'vectors'
     >>> data_kind(data=[1, 2, 3])  # 1-D sequence
-    'matrix'
+    'vectors'
     >>> data_kind(data=[[1, 2, 3], [4, 5, 6]])  # sequence of sequences
-    'matrix'
+    'vectors'
     >>> data_kind(data={"x": [1, 2, 3], "y": [4, 5, 6]})  # dictionary
-    'matrix'
+    'vectors'
     >>> data_kind(data=pd.DataFrame({"x": [1, 2, 3], "y": [4, 5, 6]}))  # pd.DataFrame
-    'matrix'
+    'vectors'
     >>> data_kind(data=pd.Series([1, 2, 3], name="x"))  # pd.Series
-    'matrix'
-
-    The "vectors" kind:
-
-    >>> data_kind(data=None)
     'vectors'
     """
     match data:
+        case None if required:  # No data provided and required=True.
+            kind = "empty"
         case str() | pathlib.PurePath():  # One file.
             kind = "file"
         case list() | tuple() if all(
@@ -312,7 +323,7 @@ def data_kind(
             # geopandas.GeoDataFrame or shapely.geometry).
             # Reference: https://gist.github.com/sgillies/2217756
             kind = "geojson"
-        case x if x is not None:  # Any not-None is considered as a matrix.
+        case np.ndarray() if data.ndim == 2:  # A 2-D numpy.ndarray object.
             kind = "matrix"
         case _:  # Fall back to "vectors" if data is None and required=True.
             kind = "vectors"

pygmt/src/legend.py

@@ -91,7 +91,7 @@ def legend(
             kwargs["F"] = box
 
     kind = data_kind(spec)
-    if kind not in {"vectors", "file", "stringio"}:  # kind="vectors" means spec is None
+    if kind not in {"empty", "file", "stringio"}:
         raise GMTInvalidInput(f"Unrecognized data type: {type(spec)}")
     if kind == "file" and is_nonstr_iter(spec):
         raise GMTInvalidInput("Only one legend specification file is allowed.")

pygmt/src/plot.py

@@ -206,7 +206,7 @@ def plot(self, data=None, x=None, y=None, size=None, direction=None, **kwargs):
 
     kind = data_kind(data)
     extra_arrays = []
-    if kind == "vectors":  # Add more columns for vectors input
+    if kind == "empty":  # Add more columns for vectors input
         # Parameters for vector styles
         if (
             kwargs.get("S") is not None

pygmt/src/plot3d.py

@@ -184,7 +184,7 @@ def plot3d(
     kind = data_kind(data)
     extra_arrays = []
 
-    if kind == "vectors":  # Add more columns for vectors input
+    if kind == "empty":  # Add more columns for vectors input
         # Parameters for vector styles
         if (
             kwargs.get("S") is not None

pygmt/src/text.py

@@ -196,7 +196,7 @@ def text_(  # noqa: PLR0912
         raise GMTInvalidInput("'text' can't be None or array when 'position' is given.")
     if textfiles is not None and text is not None:
         raise GMTInvalidInput("'text' can't be specified when 'textfiles' is given.")
-    if kind == "vectors" and text is None:
+    if kind == "empty" and text is None:
         raise GMTInvalidInput("Must provide text with x/y pairs.")
 
     # Arguments that can accept arrays.
@@ -220,7 +220,7 @@ def text_(  # noqa: PLR0912
 
     extra_arrays = []
     confdict = {}
-    if kind == "vectors":
+    if kind == "empty":
         for arg, flag, name in array_args:
             if is_nonstr_iter(arg):
                 kwargs["F"] += flag

pygmt/src/x2sys_cross.py

@@ -195,7 +195,7 @@ def x2sys_cross(
         match data_kind(track):
             case "file":
                 file_contexts.append(contextlib.nullcontext(track))
-            case "matrix":
+            case "vectors":
                 # find suffix (-E) of trackfiles used (e.g. xyz, csv, etc) from
                 # $X2SYS_HOME/TAGNAME/TAGNAME.tag file
                 tagfile = Path(