Source code for pyart.map.gates_to_grid

"""
Generate a Cartesian grid by mapping from radar gates onto the grid.

"""

import warnings

import numpy as np

from ..core.radar import Radar
from ..core.transforms import geographic_to_cartesian
from ..filters import GateFilter, moment_based_gate_filter

from ._gate_to_grid_map import GateToGridMapper
from ._gate_to_grid_map import RoIFunction, ConstantRoI, DistBeamRoI, DistRoI


def map_gates_to_grid(
        radars, grid_shape, grid_limits, grid_origin=None,
        grid_origin_alt=None, grid_projection=None,
        fields=None, gatefilters=False, map_roi=True,
        weighting_function='Barnes2', toa=17000.0, roi_func='dist_beam',
        constant_roi=None, z_factor=0.05, xy_factor=0.02, min_radius=500.0,
        h_factor=1.0, nb=1.5, bsp=1.0, **kwargs):
    """
    Map gates from one or more radars to a Cartesian grid.

    Generate a Cartesian grid of points for the requested fields from the
    collected points from one or more radars. For each radar gate that is not
    filtered a radius of influence is calculated. The weighted field values
    for that gate are added to all grid points within that radius. This
    routine scaled linearly with the number of radar gates and the effective
    grid size.

    Parameters not defined below are identical to those in
    :py:func:`map_to_grid`.

    Parameters
    ----------
    roi_func : str or RoIFunction
        Radius of influence function. A function which takes an
        z, y, x grid location, in meters, and returns a radius (in meters)
        within which all collected points will be included in the weighting
        for that grid points. Examples can be found in the
        Typically following strings can use to specify a built in
        radius of influence function:

            * constant: constant radius of influence.
            * dist: radius grows with the distance from each radar.
            * dist_beam: radius grows with the distance from each radar
              and parameter are based of virtual beam sizes.

        A custom RoIFunction can be defined using the RoIFunction class
        and defining a get_roi method which returns the radius. For efficient
        mapping this class should be implemented in Cython.

    Returns
    -------
    grids : dict
        Dictionary of mapped fields. The keys of the dictionary are given by
        parameter fields. Each elements is a `grid_size` float64 array
        containing the interpolated grid for that field.

    See Also
    --------
    grid_from_radars : Map to a grid and return a Grid object
    map_to_grid : Create grid by finding the radius of influence around each
                  grid point.

    """
    # make a tuple if passed a radar object as the first argument
    if isinstance(radars, Radar):
        radars = (radars, )

    if len(radars) == 0:
        raise ValueError('Length of radars tuple cannot be zero')

    skip_transform = False
    if len(radars) == 1 and grid_origin_alt is None and grid_origin is None:
        skip_transform = True

    if grid_origin_alt is None:
        try:
            grid_origin_alt = float(radars[0].altitude['data'])
        except TypeError:
            grid_origin_alt = np.mean(radars[0].altitude['data'])

    gatefilters = _parse_gatefilters(gatefilters, radars)
    cy_weighting_function = _detemine_cy_weighting_func(weighting_function)
    projparams = _find_projparams(grid_origin, radars, grid_projection)
    fields = _determine_fields(fields, radars)
    grid_starts, grid_steps = _find_grid_params(grid_shape, grid_limits)
    offsets = _find_offsets(radars, projparams, grid_origin_alt)
    roi_func = _parse_roi_func(roi_func, constant_roi, z_factor, xy_factor,
                               min_radius, h_factor, nb, bsp, offsets)

    # prepare grid storage arrays
    nfields = len(fields)
    grid_sum = np.zeros(grid_shape + (nfields, ), dtype=np.float32)
    grid_wsum = np.zeros(grid_shape + (nfields, ), dtype=np.float32)
    gatemapper = GateToGridMapper(
        grid_shape, grid_starts, grid_steps, grid_sum, grid_wsum)

    # project gates from each radar onto the grid
    for radar, gatefilter in zip(radars, gatefilters):

        # Copy the field data and masks.
        # TODO method that does not copy field data into new array
        if nfields == 0:
            raise ValueError(
                'There are 0 fields in the radar object to interpolate!')
        shape = (radar.nrays, radar.ngates, nfields)
        field_data = np.empty(shape, dtype='float32')
        field_mask = np.empty(shape, dtype='uint8')
        for i, field in enumerate(fields):
            fdata = radar.fields[field]['data']
            field_data[:, :, i] = np.ma.getdata(fdata)
            field_mask[:, :, i] = np.ma.getmaskarray(fdata)

        # find excluded gates from the gatefilter
        if gatefilter is False:
            gatefilter = GateFilter(radar)  # include all gates
        elif gatefilter is None:
            gatefilter = moment_based_gate_filter(radar, **kwargs)
        excluded_gates = gatefilter.gate_excluded.astype('uint8')

        # calculate gate locations relative to the grid origin
        if skip_transform:
            # single radar, grid centered at radar location
            gate_x = radar.gate_x['data']
            gate_y = radar.gate_y['data']
        else:
            gate_x, gate_y = geographic_to_cartesian(
                radar.gate_longitude['data'], radar.gate_latitude['data'],
                projparams)
        gate_z = radar.gate_altitude['data'] - grid_origin_alt

        # map the gates onto the grid
        gatemapper.map_gates_to_grid(
            radar.ngates, radar.nrays, gate_z.astype('float32'),
            gate_y.astype('float32'), gate_x.astype('float32'),
            field_data, field_mask, excluded_gates,
            toa, roi_func, cy_weighting_function)

    # create and return the grid dictionary
    mweight = np.ma.masked_equal(grid_wsum, 0)
    msum = np.ma.masked_array(grid_sum, mweight.mask)
    grids = dict(
        [(f, msum[..., i] / mweight[..., i]) for i, f in enumerate(fields)])
    if map_roi:
        roi_array = np.empty(grid_shape, dtype=np.float32)
        gatemapper.find_roi_for_grid(roi_array, roi_func)
        grids['ROI'] = roi_array
    return grids


def _detemine_cy_weighting_func(weighting_function):
    """ Determine cython weight function value. """
    if weighting_function.upper() == 'BARNES2':
        cy_weighting_function = 3
    elif weighting_function.upper() == 'NEAREST':
        cy_weighting_function = 2
    elif weighting_function.upper() == 'CRESSMAN':
        cy_weighting_function = 1
    elif weighting_function.upper() == 'BARNES':
        warnings.warn("Barnes weighting function is deprecated."
                      " Please use Barnes 2 to be consistent with"
                      " Pauley and Wu 1990. Default will be switched"
                      " to Barnes2 on June 1st.", DeprecationWarning)
        cy_weighting_function = 0
    else:
        raise ValueError('unknown weighting_function')
    return cy_weighting_function


def _find_projparams(grid_origin, radars, grid_projection):
    """ Determine the projection parameter. """

    # parse grid_origin
    if grid_origin is None:
        try:
            lat = float(radars[0].latitude['data'])
            lon = float(radars[0].longitude['data'])
        except TypeError:
            lat = np.mean(radars[0].latitude['data'])
            lon = np.mean(radars[0].longitude['data'])
        grid_origin = (lat, lon)

    grid_origin_lat, grid_origin_lon = grid_origin

    # parse grid_projection
    if grid_projection is None:
        grid_projection = {
            'proj': 'pyart_aeqd', '_include_lon_0_lat_0': True}
    projparams = grid_projection.copy()
    if projparams.pop('_include_lon_0_lat_0', False):
        projparams['lon_0'] = grid_origin_lon
        projparams['lat_0'] = grid_origin_lat
    return projparams


def _parse_gatefilters(gatefilters, radars):
    """ Parse the gatefilters parameter. """
    if isinstance(gatefilters, GateFilter):
        gatefilters = (gatefilters, )  # make tuple if single filter passed
    if gatefilters is False:
        gatefilters = (False, ) * len(radars)
    if gatefilters is None:
        gatefilters = (None, ) * len(radars)
    if len(gatefilters) != len(radars):
        raise ValueError('Length of gatefilters must match length of radars')
    return gatefilters


def _determine_fields(fields, radars):
    """ Determine which field should be mapped to the grid. """
    if fields is None:
        fields = set(radars[0].fields.keys())
        for radar in radars[1:]:
            fields = fields.intersection(radar.fields.keys())
        fields = list(fields)
    return fields


def _find_offsets(radars, projparams, grid_origin_alt):
    """ Find offset between radars and grid origin. """
    # loop over the radars finding offsets from the origin
    offsets = [] # offsets from the grid origin, in meters, for each radar
    for radar in radars:
        x_disp, y_disp = geographic_to_cartesian(
            radar.longitude['data'], radar.latitude['data'], projparams)
        try:
            z_disp = float(radar.altitude['data']) - grid_origin_alt
            offsets.append((z_disp, float(y_disp), float(x_disp)))
        except TypeError:
            z_disp = np.mean(radar.altitude['data']) - grid_origin_alt
            offsets.append((z_disp, np.mean(y_disp), np.mean(x_disp)))
    return offsets


def _find_grid_params(grid_shape, grid_limits):
    """ Find the starting points and step size of the grid. """
    nz, ny, nx = grid_shape
    zr, yr, xr = grid_limits
    z_start, z_stop = zr
    y_start, y_stop = yr
    x_start, x_stop = xr

    if nz == 1:
        z_step = 0.
    else:
        z_step = (z_stop - z_start) / (nz - 1.)
    if ny == 1:
        y_step = 0.
    else:
        y_step = (y_stop - y_start) / (ny - 1.)
    if nx == 1:
        x_step = 0.
    else:
        x_step = (x_stop - x_start) / (nx - 1.)

    grid_starts = (z_start, y_start, x_start)
    grid_steps = (z_step, y_step, x_step)
    return grid_starts, grid_steps


def _parse_roi_func(roi_func, constant_roi, z_factor, xy_factor, min_radius,
                    h_factor, nb, bsp, offsets):
    """ Return the Radius of influence object. """
    if not isinstance(roi_func, RoIFunction):
        if constant_roi is not None:
            roi_func = 'constant'
        else:
            constant_roi = 500.0
        if roi_func == 'constant':
            roi_func = ConstantRoI(constant_roi)
        elif roi_func == 'dist':
            roi_func = DistRoI(z_factor, xy_factor, min_radius, offsets)
        elif roi_func == 'dist_beam':
            roi_func = DistBeamRoI(h_factor, nb, bsp, min_radius, offsets)
        else:
            raise ValueError('unknown roi_func: %s' % roi_func)
    return roi_func