Python tables.open_file函数代码示例

    def test_maximum_overview_size(self):

        filename = make_temp_dir('maxisze.hdf5')

        env = Environment(trajectory='Testmigrate', filename=filename,

                          log_config=get_log_config(), add_time=True)

        traj = env.v_trajectory
        for irun in range(pypetconstants.HDF5_MAX_OVERVIEW_TABLE_LENGTH):
            traj.f_add_parameter('f%d.x' % irun, 5)

        traj.f_store()


        store = pt.open_file(filename, mode='r+')
        table = store.root._f_get_child(traj.v_name).overview.parameters_overview
        self.assertEquals(table.nrows, pypetconstants.HDF5_MAX_OVERVIEW_TABLE_LENGTH)
        store.close()

        for irun in range(pypetconstants.HDF5_MAX_OVERVIEW_TABLE_LENGTH,
                  2*pypetconstants.HDF5_MAX_OVERVIEW_TABLE_LENGTH):
            traj.f_add_parameter('f%d.x' % irun, 5)

        traj.f_store()

        store = pt.open_file(filename, mode='r+')
        table = store.root._f_get_child(traj.v_name).overview.parameters_overview
        self.assertEquals(table.nrows, pypetconstants.HDF5_MAX_OVERVIEW_TABLE_LENGTH)
        store.close()

        env.f_disable_logging()

    def test02_copy(self):
        """Checking (X)Array.copy() method ('numetic' flavor)"""

        srcfile = self._testFilename("oldflavor_numeric.h5")
        tmpfile = tempfile.mktemp(".h5")
        shutil.copy(srcfile, tmpfile)
        try:
            # Open the HDF5 with old numeric flavor
            with tables.open_file(tmpfile, "r+") as h5file:
                # Copy to another location
                self.assertWarns(FlavorWarning,
                                 h5file.root.array1.copy, '/', 'array1copy')
                h5file.root.array2.copy('/', 'array2copy')
                h5file.root.carray1.copy('/', 'carray1copy')
                h5file.root.carray2.copy('/', 'carray2copy')
                h5file.root.vlarray1.copy('/', 'vlarray1copy')
                h5file.root.vlarray2.copy('/', 'vlarray2copy')

                if self.close:
                    h5file.close()
                    h5file = tables.open_file(tmpfile)
                else:
                    h5file.flush()

                # Assert other properties in array
                self.assertEqual(h5file.root.array1copy.flavor, 'numeric')
                self.assertEqual(h5file.root.array2copy.flavor, 'python')
                self.assertEqual(h5file.root.carray1copy.flavor, 'numeric')
                self.assertEqual(h5file.root.carray2copy.flavor, 'python')
                self.assertEqual(h5file.root.vlarray1copy.flavor, 'numeric')
                self.assertEqual(h5file.root.vlarray2copy.flavor, 'python')
        finally:
            os.remove(tmpfile)

def validate_results(test, expected_path, actual_path):
    """Validate results by comparing in and output HDF5 files

    :param test: instance of the TestCase.
    :param expected_path: path to the reference data.
    :param actual_path: path to the output from the test.

    """
    with tables.open_file(expected_path, 'r') as expected_file, \
            tables.open_file(actual_path, 'r') as actual_file:
        for expected_node in expected_file.walk_nodes('/', 'Leaf'):
            try:
                actual_node = actual_file.get_node(expected_node._v_pathname)
            except tables.NoSuchNodeError:
                test.fail("Node '%s' does not exist in datafile" %
                          expected_node._v_pathname)
            if type(expected_node) is tables.table.Table:
                validate_tables(test, expected_node, actual_node)
            elif type(expected_node) is tables.vlarray.VLArray:
                validate_vlarrays(test, expected_node, actual_node)
            elif type(expected_node) is tables.array.Array:
                validate_arrays(test, expected_node, actual_node)
            else:
                raise NotImplementedError
            validate_attributes(test, expected_node, actual_node)
        validate_attributes(test, expected_file.root, actual_file.root)

    def test02_CompareTable(self):
        "Comparing written time data with read data in a Table."

        wtime = 1234567890.123456

        # Create test Table with data.
        h5file = tables.open_file(
            self.h5fname, 'w', title="Test for comparing Time tables")
        tbl = h5file.create_table('/', 'test', self.MyTimeRow)
        row = tbl.row
        row['t32col'] = int(wtime)
        row['t64col'] = (wtime, wtime)
        row.append()
        h5file.close()

        # Check the written data.
        h5file = tables.open_file(self.h5fname)
        recarr = h5file.root.test.read(0)
        h5file.close()

        self.assertEqual(recarr['t32col'][0], int(wtime),
                         "Stored and retrieved values do not match.")

        comp = (recarr['t64col'][0] == numpy.array((wtime, wtime)))
        self.assertTrue(numpy.alltrue(comp),
                        "Stored and retrieved values do not match.")

 def test_2D_multiphase(self):
     # RELOAD MODULES
     self.reload_modules()
     pnList = [(twp_navier_stokes_p, twp_navier_stokes_n),
               (clsvof_p,               clsvof_n)]
     self.so = multiphase_so
     pList=[]
     nList=[]
     sList=[]
     for (pModule,nModule) in pnList:
         pList.append(pModule)
         if pList[-1].name == None:
             pList[-1].name = pModule.__name__
         nList.append(nModule)
     for i in range(len(pnList)):
         sList.append(default_s)
     self.so.name += "_2D_falling_bubble"
     # NUMERICAL SOLUTION #
     ns = proteus.NumericalSolution.NS_base(self.so,
                                            pList,
                                            nList,
                                            sList,
                                            opts)
     ns.calculateSolution('2D_falling_bubble')
     # COMPARE VS SAVED FILES #
     expected_path = 'comparison_files/multiphase_2D_falling_bubble.h5'
     expected = tables.open_file(os.path.join(self._scriptdir,expected_path))
     actual = tables.open_file('multiphase_2D_falling_bubble.h5','r')
     assert np.allclose(expected.root.phi_t2,actual.root.phi_t2,atol=1e-10)
     expected.close()
     actual.close()

def main(argv):
    args = parse_args(argv[1:])

    fileout = os.path.abspath(args.output)

    start = time()
    for fin in args.inputs:
        filein = os.path.abspath(fin)
        print 'Concatenating %s' % filein
        if not os.path.exists(fileout):
            copyfile(filein, fileout)
        else:
            # Can't use HdfStorage.readCoordinates because it needs an
            # Ice.Communicator object, so there's no point using the
            # OMERO.tables interface
            tout = tables.open_file(fileout, 'r+')
            tin = tables.open_file(filein, 'r')
            nrows = tin.root.OME.Measurements.nrows

            for a in range(0, nrows, ROW_CHUNK):
                b = min(nrows, a + ROW_CHUNK)
                print '\tRows %d:%d' % (a, b)
                rows = tin.root.OME.Measurements.read_coordinates(range(a, b))
                tout.root.OME.Measurements.append(rows)

            tin.close()
            tout.close()
        print '\tCumulative time: %d seconds' % (time() - start)

    print 'Done'

 def test_case_2(self):
     # Set parameters for this test
     parameters.ct.test_case=2
     # RELOAD MODULES
     self.reload_modules()
     pnList = [(clsvof_p, clsvof_n)]
     self.so = default_so
     self.so.tnList = clsvof_n.tnList
     pList=[]
     nList=[]
     sList=[]
     for (pModule,nModule) in pnList:
         pList.append(pModule)
         if pList[-1].name == None:
             pList[-1].name = pModule.__name__
         nList.append(nModule)
     for i in range(len(pnList)):
         sList.append(default_s)
     self.so.name = "clsvof_test_case_2"
     # NUMERICAL SOLUTION #
     ns = proteus.NumericalSolution.NS_base(self.so,
                                            pList,
                                            nList,
                                            sList,
                                            opts)
     ns.calculateSolution('test_case_2')
     # COMPARE VS SAVED FILES #
     expected_path = 'comparison_files/clsvof_test_case_2.h5'
     expected = tables.open_file(os.path.join(self._scriptdir,expected_path))
     actual = tables.open_file('clsvof_test_case_2.h5','r')
     assert np.allclose(expected.root.u_t2,actual.root.u_t2,atol=1e-10)
     expected.close()
     actual.close()

 def __init__( self,
               fnContigLengths,
               fnWssd,
               overwrite,
               openMode,
               groupsToCheck=[],
               compression=False ):
     
     self.compress = compression
     
     assert os.path.exists(fnContigLengths)
     if openMode=='r':
         assert not overwrite
         assert os.path.exists(fnWssd), fnWssd
     
     debug_output('WssdBase: reading contig lengths from file %s'%fnContigLengths)        
     
     self.mContigNameLen = {}
     for l in open(fnContigLengths,'r'):
         l=l.replace('\n','').split('\t')
         self.mContigNameLen[l[0]]=int(l[1])
     
     debug_output('WSSD space: %d contigs totaling %d bp'%( len(self.mContigNameLen), sum(self.mContigNameLen.values()) ))
     
     if overwrite or not os.path.exists(fnWssd): 
         self.tbl = tables.open_file( fnWssd, 'w' )
     else:
         if openMode=='r':
             self.tbl = tables.open_file( fnWssd, 'r' )
         else:
             self.tbl = tables.open_file( fnWssd, 'a' )

 def process(self, rows_slice):
                 
     with Worker.hdf5_lock:
         with tables.open_file(self.hdf5_file, 'r+') as fileh:
             T = fileh.get_node(self.path + '/temporaries')
             tmp = T[rows_slice, ...]
         
     ind = np.arange(0, rows_slice.stop - rows_slice.start)    
     
     # tmp = - A_new    
     tmp -= self.rows_sum
     diag_A = tmp[ind, rows_slice.start + ind].copy()
     np.clip(tmp, 0, np.inf, tmp)
     tmp[ind, rows_slice.start + ind] = diag_A
     
     Worker.hdf5_lock.acquire()
     
     with tables.open_file(self.hdf5_file, 'r+') as fileh:
         A = fileh.get_node(self.path + '/availabilities')
         a = A[rows_slice, ...]
         
     Worker.hdf5_lock.release()
         
     # yet more damping
     a = a * self.damping - tmp * (1 - self.damping)
     
     with Worker.hdf5_lock:
         with tables.open_file(self.hdf5_file, 'r+') as fileh:
             A = fileh.get_node(self.path + '/availabilities')
             T = fileh.get_node(self.path + '/temporaries')
             
             A[rows_slice, ...] = a
             T[rows_slice, ...] = tmp
             
     del a, tmp

def create_synth(kind, prec):

    prefix_orig = "cellzome/cellzome-"
    iname = dirname + prefix_orig + "none-" + prec + ".h5"
    f = tb.open_file(iname, "r")

    if prec == "single":
        type_ = tb.Float32Atom()
    else:
        type_ = tb.Float64Atom()

    prefix = "synth/synth-"
    for clevel in range(10):
        oname = "%s/%s-%s%d-%s.h5" % (dirname, prefix, kind, clevel, prec)
        # print "creating...", iname
        f2 = tb.open_file(oname, "w")

        if kind in ["none", "numpy"]:
            filters = None
        else:
            filters = tb.Filters(complib=kind, complevel=clevel, shuffle=shuffle)

        for name in ["maxarea", "mascotscore"]:
            col = f.get_node("/", name)
            r = f2.create_carray("/", name, type_, col.shape, filters=filters)
            if name == "maxarea":
                r[:] = np.arange(col.nrows, dtype=type_.dtype)
            else:
                r[:] = np.arange(col.nrows, 0, dtype=type_.dtype)

        f2.close()
        if clevel == 0:
            size = 1.5 * float(os.stat(oname)[6])
    f.close()
    return size

 def open(self, mode, ncols=1, nrows=1, xll=0, yll=0, cellsize=1, nodatavalue=-9999.0,
     dataset_name="dummy", group_prefix="row", table_prefix="col", index_format="04i", variables=[], units=[]):
     # Initialise
     fpath = os.path.join(self.folder, self.name);
     if (mode[0] == 'w'):
         # Open the file
         self.__datafile = tables.open_file(fpath, 'w');
         
         # Assign the data attributes 
         self.ncols = ncols;
         self.nrows = nrows;                    
         self.xll = xll;
         self.yll = yll;
         self.cellsize = cellsize;
         self.nodatavalue = nodatavalue;
         self.dataset_name = dataset_name;
         self.group_prefix = group_prefix;
         self.table_prefix = table_prefix;
         self.index_format = index_format;
         self.variables = variables;
         self.units = units;
         self.writeheader();
     else: 
         # If file does not exist, then ...
         if os.path.exists(fpath):
             # Retrieve the data attributes from the header file
             self.readheader();
             GridEnvelope2D.__init__(self, self.ncols, self.nrows, self.xll, self.yll, self.cellsize, self.cellsize);
             self.__datafile = tables.open_file(fpath, 'r');
             return True;
         else: return False;   

def h5_apply_func(input_path, output_path, node_func):
    """
    Apply node_func to all nodes of input_path and store the result in
    output_path

    Parameters
    ----------
    input_path : str
        path to .h5 input file
    output_path : str
        path to .h5 output file
    node_func : function
        function that will be applied to all nodes
        func(node, new_parent) -> new_node
        new_node must be node if node must be copied
                         None if node must not be copied
                         another Node if node must not be copied (was already
                                      handled/copied/modified by func)
    """
    with tables.open_file(input_path) as input_file, \
            tables.open_file(output_path, mode="w") as output_file:
        for node in input_file.walk_nodes(classname='Leaf'):
            if node is not input_file.root:
                print(node._v_pathname, "...", end=' ')
                parent_path = node._v_parent._v_pathname
                if parent_path in output_file:
                    new_parent = output_file.get_node(parent_path)
                else:
                    new_parent = output_file._create_path(parent_path)
                new_node = node_func(node, new_parent)
                if new_node is node:
                    print("copying (without modifications) ...", end=' ')
                    node._f_copy(new_parent)
                print("done.")

    def __init__(self, cfg):
        self.cfg = cfg
        self.path = os.path.join(self.cfg.subsets_path, 'data.db')
        self.results = None
        if os.path.exists(self.path):
            try:
                self.h5 = tables.open_file(self.path, 'a')
                self.results = self.h5.root.results
            except:
                # If anything fails, we just create a new database...
                log.warning("""Failed to open existing database at %s, or
                database is corrupted. Creating a new one""", self.path)
                self.results = None

        # Something went wrong!
        if not self.results:
            try:
                # Try closing this, just in case
                self.h5.close()
            except:
                pass

            # Compression is good -- and faster, according to the pytables docs...
            f = tables.Filters(complib='blosc', complevel=5)
            self.h5 = tables.open_file(self.path, 'w', filters=f)
            self.results = self.h5.create_table(
                '/', 'results', cfg.data_layout.data_type)
            self.results.cols.subset_id.create_csindex()

        assert isinstance(self.results, tables.Table)
        assert self.results.indexed

def ptconcat(output_file, input_files, overwrite=False):
    """Concatenate HDF5 Files"""
    filt = tb.Filters(
        complevel=5, shuffle=True, fletcher32=True, complib='zlib'
    )
    out_tabs = {}
    dt_file = input_files[0]
    log.info("Reading data struct '%s'..." % dt_file)
    h5struc = tb.open_file(dt_file, 'r')
    log.info("Opening output file '%s'..." % output_file)
    if overwrite:
        outmode = 'w'
    else:
        outmode = 'a'
    h5out = tb.open_file(output_file, outmode)

    for node in h5struc.walk_nodes('/', classname='Table'):
        path = node._v_pathname
        log.debug(path)
        dtype = node.dtype
        p, n = os.path.split(path)
        out_tabs[path] = h5out.create_table(
            p, n, description=dtype, filters=filt, createparents=True
        )
    h5struc.close()
    for fname in input_files:
        log.info('Reading %s...' % fname)
        h5 = tb.open_file(fname)
        for path, out in out_tabs.items():
            tab = h5.get_node(path)
            out.append(tab[:])
        h5.close()
    h5out.close()

def test_estimator_pytables():
    m1 = MyEstimator(a=1, b='a', c=None, d=False, e=np.zeros(3)).fit(None)

    f = tables.open_file(fn, 'w')
    m1.to_pytables(f.root)
    f.close()

    g = tables.open_file(fn)
    m2 = MyEstimator.from_pytables(g.root.MyEstimator)

    print m1.__dict__
    print m2.__dict__

    for key, value in m1.get_params().iteritems():
        if any(isinstance(value, t) for t in [int, float, str]):
            assert value == getattr(m2, key, object())
        else:
            eq(value, getattr(m2, key, object()), err_msg='error on param key=%s' % key)

    for key in m1._get_estimate_names():
        value = getattr(m1, key)
        if any(isinstance(value, t) for t in [int, float, str]):
            assert value == getattr(m2, key, object())
        else:
            eq(value, getattr(m2, key, object()), err_msg='error on estimate key=%s' % key)

    g.close()