tst_var_get_vars.py

#
# Copyright (C) 2024, Northwestern University and Argonne National Laboratory
# See COPYRIGHT notice in top-level directory.
#

"""
   This example program is intended to illustrate the use of the pnetCDF python API.The
   program runs in blocking mode and read an subsampled array of values from a netCDF
   variable of an opened netCDF file using iget_var method of `Variable` class. The
   library will internally invoke ncmpi_get_vars in C.
"""
import pnetcdf
from numpy.random import seed, randint
from numpy.testing import assert_array_equal, assert_equal, assert_array_almost_equal
import tempfile, unittest, os, random, sys
import numpy as np
from mpi4py import MPI
from utils import validate_nc_file
import io

seed(0)
file_formats = ['NC_64BIT_DATA', 'NC_64BIT_OFFSET', None]
file_name = "tst_var_get_vars.nc"


comm = MPI.COMM_WORLD
rank = comm.Get_rank()
size = comm.Get_size()
xdim=9; ydim=10; zdim=size*10
# initial values for netCDF variable
data = randint(0,10, size=(xdim,ydim,zdim)).astype('i4')
# generate reference arrayes for testing
dataref = []
for i in range(size):
    dataref.append(data[3:4:1,0:6:2,i*10:(i+1)*10:2])

class VariablesTestCase(unittest.TestCase):

    def setUp(self):
        if (len(sys.argv) == 2) and os.path.isdir(sys.argv[1]):
            self.file_path = os.path.join(sys.argv[1], file_name)
        else:
            self.file_path = file_name
        self._file_format = file_formats.pop(0)
        f = pnetcdf.File(filename=self.file_path, mode = 'w', format=self._file_format, comm=comm, info=None)
        f.def_dim('x',xdim)
        f.def_dim('xu',-1)
        f.def_dim('y',ydim)
        f.def_dim('z',zdim)

        v1_u = f.def_var('data1u', pnetcdf.NC_INT, ('xu','y','z'))

        #initialize variable values
        f.enddef()
        v1_u[:] = data
        f.close()
        assert validate_nc_file(os.environ.get('PNETCDF_DIR'), self.file_path) == 0 if os.environ.get('PNETCDF_DIR') is not None else True


    def runTest(self):
        """testing variable get_vars method for CDF-5/CDF-2/CDF-1 file format"""

        f = pnetcdf.File(self.file_path, 'r')
        # equivalent code to the following using indexer syntax: v1_data = v1[3:4,0:6:2,10*rank:10*(rank+1):2]
        starts = np.array([3,0,10*rank])
        counts = np.array([1,3,5])
        strides = np.array([1,2,2])
        # test collective i/o get_var
        v1 = f.variables['data1u']
        buff = np.empty(tuple(counts), v1.dtype)
        # all processes read the designated slices of the variable using collective i/o
        buff = np.empty(tuple(counts), v1.dtype)
        v1.get_var_all(buff, start = starts, count = counts, stride = strides)
        # compare returned numpy array against reference array
        assert_array_equal(buff, dataref[rank])
        # test independent i/o get_var
        f.begin_indep()
        if rank < 2:
            # mpi process rank 0 and rank 1 respectively read the assigned slice of the variable using independent i/o
            buff = np.empty(tuple(counts), v1.dtype)
            v1.get_var(buff, start = starts, count = counts, stride = strides)
            # compare returned numpy array against reference array
            assert_array_equal(buff, dataref[rank])
        f.close()

    def tearDown(self):
        # remove the temporary files if test file directory not specified
        comm.Barrier()
        if (rank == 0) and not((len(sys.argv) == 2) and os.path.isdir(sys.argv[1])):
            os.remove(self.file_path)

if __name__ == '__main__':
    suite = unittest.TestSuite()
    for i in range(len(file_formats)):
        suite.addTest(VariablesTestCase())
    runner = unittest.TextTestRunner()
    output = io.StringIO()
    runner = unittest.TextTestRunner(stream=output)
    result = runner.run(suite)
    if not result.wasSuccessful():
        print(output.getvalue())
        sys.exit(1)