Remove mlab.griddata

dstansby · dstansby · commit 66eb5044a135 · 2018-09-21T11:35:04.000+01:00
diff --git a/doc/api/next_api_changes/2018-09-18-DS.rst b/doc/api/next_api_changes/2018-09-18-DS.rst
@@ -21,4 +21,5 @@ in Matplotlib 2.2 has been removed. See below for a list:
 - `mlab.get_sparse_matrix`
 - `mlab.dist` (use numpy.hypot instead)
 - `mlab.dist_point_to_segment`
+- `mlab.griddata` (use scipy.interpolate.griddata)
 - `mlab.donothing_callback`
diff --git a/lib/matplotlib/mlab.py b/lib/matplotlib/mlab.py
@@ -2556,132 +2556,6 @@ def newfunc(val, mask, mval):
         fh.close()
 
 
-@cbook.deprecated('2.2', alternative='scipy.interpolate.griddata')
-def griddata(x, y, z, xi, yi, interp='nn'):
-    """
-    Interpolates from a nonuniformly spaced grid to some other grid.
-
-    Fits a surface of the form z = f(`x`, `y`) to the data in the
-    (usually) nonuniformly spaced vectors (`x`, `y`, `z`), then
-    interpolates this surface at the points specified by
-    (`xi`, `yi`) to produce `zi`.
-
-    Parameters
-    ----------
-    x, y, z : 1d array_like
-        Coordinates of grid points to interpolate from.
-    xi, yi : 1d or 2d array_like
-        Coordinates of grid points to interpolate to.
-    interp : string key from {'nn', 'linear'}
-        Interpolation algorithm, either 'nn' for natural neighbor, or
-        'linear' for linear interpolation.
-
-    Returns
-    -------
-    2d float array
-        Array of values interpolated at (`xi`, `yi`) points.  Array
-        will be masked is any of (`xi`, `yi`) are outside the convex
-        hull of (`x`, `y`).
-
-    Notes
-    -----
-    If `interp` is 'nn' (the default), uses natural neighbor
-    interpolation based on Delaunay triangulation.  This option is
-    only available if the mpl_toolkits.natgrid module is installed.
-    This can be downloaded from https://github.com/matplotlib/natgrid.
-    The (`xi`, `yi`) grid must be regular and monotonically increasing
-    in this case.
-
-    If `interp` is 'linear', linear interpolation is used via
-    matplotlib.tri.LinearTriInterpolator.
-
-    Instead of using `griddata`, more flexible functionality and other
-    interpolation options are available using a
-    matplotlib.tri.Triangulation and a matplotlib.tri.TriInterpolator.
-    """
-    # Check input arguments.
-    x = np.asanyarray(x, dtype=np.float64)
-    y = np.asanyarray(y, dtype=np.float64)
-    z = np.asanyarray(z, dtype=np.float64)
-    if x.shape != y.shape or x.shape != z.shape or x.ndim != 1:
-        raise ValueError("x, y and z must be equal-length 1-D arrays")
-
-    xi = np.asanyarray(xi, dtype=np.float64)
-    yi = np.asanyarray(yi, dtype=np.float64)
-    if xi.ndim != yi.ndim:
-        raise ValueError("xi and yi must be arrays with the same number of "
-                         "dimensions (1 or 2)")
-    if xi.ndim == 2 and xi.shape != yi.shape:
-        raise ValueError("if xi and yi are 2D arrays, they must have the same "
-                         "shape")
-    if xi.ndim == 1:
-        xi, yi = np.meshgrid(xi, yi)
-
-    if interp == 'nn':
-        use_nn_interpolation = True
-    elif interp == 'linear':
-        use_nn_interpolation = False
-    else:
-        raise ValueError("interp keyword must be one of 'linear' (for linear "
-                         "interpolation) or 'nn' (for natural neighbor "
-                         "interpolation).  Default is 'nn'.")
-
-    # Remove masked points.
-    mask = np.ma.getmask(z)
-    if mask is not np.ma.nomask:
-        x = x.compress(~mask)
-        y = y.compress(~mask)
-        z = z.compressed()
-
-    if use_nn_interpolation:
-        try:
-            from mpl_toolkits.natgrid import _natgrid
-        except ImportError:
-            raise RuntimeError(
-                "To use interp='nn' (Natural Neighbor interpolation) in "
-                "griddata, natgrid must be installed. Either install it "
-                "from http://github.com/matplotlib/natgrid or use "
-                "interp='linear' instead.")
-
-        if xi.ndim == 2:
-            # natgrid expects 1D xi and yi arrays.
-            xi = xi[0, :]
-            yi = yi[:, 0]
-
-        # Override default natgrid internal parameters.
-        _natgrid.seti(b'ext', 0)
-        _natgrid.setr(b'nul', np.nan)
-
-        if np.min(np.diff(xi)) < 0 or np.min(np.diff(yi)) < 0:
-            raise ValueError("Output grid defined by xi,yi must be monotone "
-                             "increasing")
-
-        # Allocate array for output (buffer will be overwritten by natgridd)
-        zi = np.empty((yi.shape[0], xi.shape[0]), np.float64)
-
-        # Natgrid requires each array to be contiguous rather than e.g. a view
-        # that is a non-contiguous slice of another array.  Use numpy.require
-        # to deal with this, which will copy if necessary.
-        x = np.require(x, requirements=['C'])
-        y = np.require(y, requirements=['C'])
-        z = np.require(z, requirements=['C'])
-        xi = np.require(xi, requirements=['C'])
-        yi = np.require(yi, requirements=['C'])
-        _natgrid.natgridd(x, y, z, xi, yi, zi)
-
-        # Mask points on grid outside convex hull of input data.
-        if np.any(np.isnan(zi)):
-            zi = np.ma.masked_where(np.isnan(zi), zi)
-        return zi
-    else:
-        # Linear interpolation performed using a matplotlib.tri.Triangulation
-        # and a matplotlib.tri.LinearTriInterpolator.
-        from .tri import Triangulation, LinearTriInterpolator
-        triang = Triangulation(x, y)
-        interpolator = LinearTriInterpolator(triang, z)
-        return interpolator(xi, yi)
-
-
 ##################################################
 # Linear interpolation algorithms
 ##################################################
diff --git a/lib/matplotlib/pylab.py b/lib/matplotlib/pylab.py
@@ -195,7 +195,6 @@
 _Other
 
   angle     - the angle of a complex array
-  griddata  - interpolate irregularly distributed data to a regular grid
   load      - Deprecated--please use loadtxt.
   loadtxt   - load ASCII data into array.
   polyfit   - fit x, y to an n-th order polynomial
@@ -233,7 +232,7 @@
     amap, base_repr, binary_repr, csv2rec,
     demean, detrend, detrend_linear, detrend_mean, detrend_none,
     distances_along_curve, exp_safe,
-    fftsurr, frange, griddata,
+    fftsurr, frange,
     identity, inside_poly, is_closed_polygon, ispower2, isvector, l1norm,
     l2norm, log2, movavg, norm_flat,
     path_length, poly_below, poly_between,
diff --git a/lib/matplotlib/tests/test_mlab.py b/lib/matplotlib/tests/test_mlab.py
@@ -2081,81 +2081,6 @@ def test_cohere():
     assert np.isreal(np.mean(cohsq))
 
 
-def test_griddata_linear():
-    # z is a linear function of x and y.
-    def get_z(x, y):
-        return 3.0*x - y
-
-    # Passing 1D xi and yi arrays to griddata.
-    x = np.asarray([0.0, 1.0, 0.0, 1.0, 0.5])
-    y = np.asarray([0.0, 0.0, 1.0, 1.0, 0.5])
-    z = get_z(x, y)
-    xi = [0.2, 0.4, 0.6, 0.8]
-    yi = [0.1, 0.3, 0.7, 0.9]
-    with pytest.warns(MatplotlibDeprecationWarning):
-        zi = mlab.griddata(x, y, z, xi, yi, interp='linear')
-    xi, yi = np.meshgrid(xi, yi)
-    np.testing.assert_array_almost_equal(zi, get_z(xi, yi))
-
-    # Passing 2D xi and yi arrays to griddata.
-    with pytest.warns(MatplotlibDeprecationWarning):
-        zi = mlab.griddata(x, y, z, xi, yi, interp='linear')
-    np.testing.assert_array_almost_equal(zi, get_z(xi, yi))
-
-    # Masking z array.
-    z_masked = np.ma.array(z, mask=[False, False, False, True, False])
-    correct_zi_masked = np.ma.masked_where(xi + yi > 1.0, get_z(xi, yi))
-    with pytest.warns(MatplotlibDeprecationWarning):
-        zi = mlab.griddata(x, y, z_masked, xi, yi, interp='linear')
-    matest.assert_array_almost_equal(zi, correct_zi_masked)
-    np.testing.assert_array_equal(np.ma.getmask(zi),
-                                  np.ma.getmask(correct_zi_masked))
-
-
-def test_griddata_nn():
-    pytest.importorskip('mpl_toolkits.natgrid')
-
-    # z is a linear function of x and y.
-    def get_z(x, y):
-        return 3.0*x - y
-
-    # Passing 1D xi and yi arrays to griddata.
-    x = np.asarray([0.0, 1.0, 0.0, 1.0, 0.5])
-    y = np.asarray([0.0, 0.0, 1.0, 1.0, 0.5])
-    z = get_z(x, y)
-    xi = [0.2, 0.4, 0.6, 0.8]
-    yi = [0.1, 0.3, 0.7, 0.9]
-    correct_zi = [[0.49999252, 1.0999978, 1.7000030, 2.3000080],
-                  [0.29999208, 0.8999978, 1.5000029, 2.1000059],
-                  [-0.1000099, 0.4999943, 1.0999964, 1.6999979],
-                  [-0.3000128, 0.2999894, 0.8999913, 1.4999933]]
-    with pytest.warns(MatplotlibDeprecationWarning):
-        zi = mlab.griddata(x, y, z, xi, yi, interp='nn')
-    np.testing.assert_array_almost_equal(zi, correct_zi, 5)
-
-    with pytest.warns(MatplotlibDeprecationWarning):
-        # Decreasing xi or yi should raise ValueError.
-        with pytest.raises(ValueError):
-            mlab.griddata(x, y, z, xi[::-1], yi, interp='nn')
-        with pytest.raises(ValueError):
-            mlab.griddata(x, y, z, xi, yi[::-1], interp='nn')
-
-    # Passing 2D xi and yi arrays to griddata.
-    xi, yi = np.meshgrid(xi, yi)
-    with pytest.warns(MatplotlibDeprecationWarning):
-        zi = mlab.griddata(x, y, z, xi, yi, interp='nn')
-    np.testing.assert_array_almost_equal(zi, correct_zi, 5)
-
-    # Masking z array.
-    z_masked = np.ma.array(z, mask=[False, False, False, True, False])
-    correct_zi_masked = np.ma.masked_where(xi + yi > 1.0, correct_zi)
-    with pytest.warns(MatplotlibDeprecationWarning):
-        zi = mlab.griddata(x, y, z_masked, xi, yi, interp='nn')
-    np.testing.assert_array_almost_equal(zi, correct_zi_masked, 5)
-    np.testing.assert_array_equal(np.ma.getmask(zi),
-                                  np.ma.getmask(correct_zi_masked))
-
-
 #*****************************************************************
 # These Tests where taken from SCIPY with some minor modifications
 # this can be retrieved from:
