seaduck.utils

seaduck.utils#

seaduck.utils.NoneIn(lst)#

See if there is a None in the iterable object. Return a Boolean.

seaduck.utils.chg_ref_lon(x, ref_lon)#

Change the definition of 0 longitude.

Return how much east one need to go from ref_lon to x This function aims to address the confusion caused by the discontinuity in longitude.

seaduck.utils.convert_time(time)#

Convert time into seconds after 1970-01-01.

time needs to be a string or a np.datetime64 object.

seaduck.utils.create_tree(x, y, R=6371.0, leafsize=16)#

Create a cKD tree object.

Parameters:

  • x (np.ndarray) – longitude and latitude of the grid location

  • y (np.ndarray) – longitude and latitude of the grid location

  • R (float) – The radius in kilometers of the planet.

  • leafsize (int) – When to switch to brute force search.

seaduck.utils.easy_3d_cube(lon, lat, dep, tim, print_total_number=False)#

Create 4D coords for initializing Position/Particle.

seaduck.utils.find_cs_sn(thetaA, phiA, thetaB, phiB)#

Find a spherical angle OAB.

theta is the angle between the meridian crossing point A and the geodesic connecting A and B.

this function return cos and sin of theta

seaduck.utils.find_ind(array, value, peri=None, ascending=1, above=True)#

Find the index of the nearest value to the given value.

Parameters:

  • array (numpy.ndarray) – 1D numpy array to search index from

  • value (number) – The value to find nearest neighbor with

  • peri (number, optional) – The periodicity of the array. For example, 360 for longitude.

  • ascending (int, default 1) – Whether the array is in ascending order. 1 for ascending order, -1 for descending order.

  • above (boolean, default True) – If True, return the index of the largest item in array smaller than value. Otherwise, return the closest value.

seaduck.utils.find_ind_h(lons, lats, tree, h_shape)#

Use ckd tree to find the horizontal indexes,.

seaduck.utils.find_px_py(XG, YG, tp, ind, cuvwg='G')#

Find the nearest 4 corner points.

This is used in oceanparcel interpolation scheme.

seaduck.utils.find_rel(value, array, darray=None, ascending=1, above=True, peri=None, dx_right=True)#

Find the rel-coords of the 1D coords.

The backend for all find_rel functions

Parameters:

  • value (numpy.ndarray) – 1D array for the value to find rel-coords.

  • array (numpy.ndarray) – The array of potential reference levels.

  • darray (numpy.ndarray, optional) – The distances between reference levels.

  • peri (number, optional) – The periodicity of the array. For example, 360 for longitude.

  • ascending (int, default 1) – Whether the array is in ascending order. 1 for ascending order, -1 for descending order.

  • above (boolean, default True) – If True, return the index of the largest item in array smaller than value. Otherwise, return the closest value.

  • dx_right (boolean, default True) – If True, darray[i] = abs(array[i+1] - array[i])

Returns:

  • ix (numpy.ndarray) – Indexes of the reference level

  • rx (numpy.ndarray) – Non-dimensional distance to the reference level

  • dx (numpy.ndarray) – distance between the reference t level and the next one.

  • bx (numpy.ndarray) – Value of the reference level

seaduck.utils.find_rel_h_naive(lon, lat, some_x, some_y, some_dx, some_dy, CS, SN, tree)#

Find the rel-coords in the horizontal.

very similar to find_rel_time/v rx,ry,dx,dy are defined the same way for example rx = “how much to the right of the node”/”size of the cell in left-right direction” dx = “size of the cell in left-right direction”.

cs,sn is just the cos and sin of the grid orientation. It will come in handy when we transfer vectors.

seaduck.utils.find_rel_h_oceanparcel(x, y, some_x, some_y, some_dx, some_dy, CS, SN, XG, YG, tree, tp)#

Find the rel-coords using the rectilinear scheme.

seaduck.utils.find_rel_h_rectilinear(x, y, lon, lat)#

Find the rel-coords using the rectilinear scheme.

seaduck.utils.find_rel_nearest(value, ts)#

Find the rel-coords based on the find_ind_nearest method.

seaduck.utils.find_rel_periodic(value, ts, peri)#

Find the rel-coords based on the find_ind_periodic method.

seaduck.utils.find_rel_time(time, ts)#

Find the rel-coords of the temporal coords.

Parameters:

  • time (numpy.ndarray) – 1D array for the time since 1970-01-01 in seconds.

  • ts (numpy.ndarray) – The time of model time steps also in seconds.

Returns:

  • it (numpy.ndarray) – Indexes of the reference t level

  • rt (numpy.ndarray) – Non-dimensional distance to the reference t level

  • dt (numpy.ndarray) – distance between the reference t level and the next one.

seaduck.utils.find_rel_z(depth, some_z, some_dz=None, dz_above_z=True)#

Find the rel-coords of the vertical coords.

Parameters:

  • depth (numpy.ndarray) – 1D array for the depth of interest in meters. More negative means deeper.

  • some_z (numpy.ndarray) – The depth of reference depth.

  • some_dz (numpy.ndarray or None) – dz_i = abs(z_{i+1}- z_i)

  • dz_above_z (Boolean) – Whether the dz as the distance between the depth level and a shallower one(True) or a deeper one(False)

Returns:

  • iz (numpy.ndarray) – Indexes of the reference z level

  • rz (numpy.ndarray) – Non-dimensional distance to the reference z level

  • dz (numpy.ndarray) – distance between the reference z level and the next one.

seaduck.utils.find_rx_ry_naive(x, y, bx, by, cs, sn, dx, dy)#

Find the non-dimensional coords using the local cartesian scheme.

seaduck.utils.find_rx_ry_oceanparcel(x, y, px, py)#

Find the non-dimensional horizontal distance.

This is done using the oceanparcel scheme.

seaduck.utils.get_dataset(name)#

Use pooch to download datasets from cloud.

This is just for testing purposes.

Parameters:

name (string) – The name of dataset, now support “ecco”, “aviso”, “curv”, “rect”

seaduck.utils.get_key_by_value(d, value)#

Find one of the keys in a dictionary.

the key that correspond to the given value.

Parameters:

  • d (dictionaty) – dictionary to lookup key from

  • value (object) – A object that has __eq__ method.

seaduck.utils.local_to_latlon(u, v, cs, sn)#

Convert local vector to north-east.

seaduck.utils.missing_cs_sn(ds, return_xr=False)#

Fill in the CS,SN of a dataset.

seaduck.utils.parallelpointinpolygon(xs, ys, poly)#

Check if xs,ys is in the polygon, return same size boolean array.

Parameters:

  • xs (1D np.array) – the x,y locations

  • ys (1D np.array) – the x,y locations

  • poly (2D np.array) – the location of the edge of polygon, the order matters.

seaduck.utils.pointinpolygon(x, y, poly)#

Check if x,y is in the polygon.

seaduck.utils.pooch_prepare()#

Prepare for loading datasets using pooch.

seaduck.utils.process_ecco(ds)#

Add more meat to ECCO dataset after the skeleton is downloaded.

seaduck.utils.rel2latlon(rx, ry, cs, sn, dx, dy, bx, by)#

Translate the spatial rel-coords into lat-lon-dep coords.

seaduck.utils.spherical2cartesian(lat, lon, R=6371.0)#

Convert spherical coordinates to cartesian.

Parameters:

  • lat (np.array) – Spherical Y coordinate (latitude)

  • lon (np.array) – Spherical X coordinate (longitude)

  • R (scalar) – Earth radius in km If None, use geopy default

Returns:

  • x (np.array) – Cartesian x coordinate

  • y (np.array) – Cartesian y coordinate

  • z (np.array) – Cartesian z coordinate

seaduck.utils.to_180(x, peri=360)#

Convert any longitude scale to [-180,180).

seaduck.utils.weight_f_node(rx, ry)#

Assign weights to four corners.

assign weight based on the non-dimensional coords to the four corner points.