Compressor: Field compression module

Preamble

Compressor enables fields compression for arrays/pyTrees.

This module is part of Cassiopee, a free open-source pre- and post-processor for CFD simulations.

To use the module with the Compressor array interface:

import Compressor

To use the module with the CGNS/Python interface:

import Compressor.PyTree as Compressor

List of functions

– Index field compression

Compressor.deltaIndex(index, ref)

Return the delta between index and ref.

– Object serializer/compression

Compressor.pack(a[, method])

Serialize or compress a.

Compressor.unpack(a[, method])

Deserialize or decompress a.

– CGNS Zones/tree compression

Compressor.PyTree.compressCartesian(t[, …])

For Cartesian grids, replace Grid Coordinates with a UserDefined CartesianData node.

Compressor.PyTree.uncompressCartesian(t)

For Cartesian grids, recreate Grid Coordinates from compressed zones.

Compressor.PyTree.compressCellN(t)

Compress cellN on 2 bits.

Compressor.PyTree.compressCoords(t[, tol, ctype])

Compress coordinates with a relative tolerance.

Compressor.PyTree.compressFields(t[, tol, ctype])

Compress fields with a relative tolerance.

Compressor.PyTree.compressElements(t)

Compress Element connectivity.

Compressor.PyTree.uncompressAll(t)

Uncompress all compressed data.

Contents

Index field compression

Compressor.deltaIndex(a, ref)

Compress a list of indices using delta algorithm. The return Delta contains the number of added indices in a when compared to ref, the list of added indices, the number of suppressed indices, the list of suppressed indices.

Parameters

  • a (numpy of ints) – input indices

  • ref (numpy) – compared indices

Returns

list of added indices, the number of supressed indices, list of suppress indices

Return type

(numpy, int, numpy)

# - deltaIndex -
import numpy
import Compressor

# Liste des indexes de reference
indRef = numpy.array([1,2,3,4,5], dtype='int32')

# Liste des indexes a comparer a la reference
index = numpy.array([1,2,3,4], dtype='int32')

delta = Compressor.deltaIndex(index, indRef)
print(delta)

Object serialize/compression

Compressor.pack(a)

Serialize/compress a python object a. For now, this is only a general interface to pickle module.

Parameters

a (python object) – any python object

Returns

serialized stream

# - pack -
import Compressor
import Generator.PyTree as G
a = G.cart((0,0,0), (1,1,1), (1000,100,100))
b = Compressor.pack(a)
Compressor.unpack(a)

Deserialize/decompress a serialized stream b. For now, this is only a general interface to pickle module.

Parameters

a (serialized stream) – a serialized stream as produced by pack

Returns

python object

# - unpack -
import Compressor
import Generator.PyTree as G
a = G.cart((0,0,0), (1,1,1), (1000,100,100))
b = Compressor.pack(a)
c = Compressor.unpack(b)
Compressor.PyTree.compressCartesian(a)

Compress zones if they are regular Cartesian grids. Create a CartesianData node containing the 6 floats corresponding to first point and steps in 3 directions.

Exists also as an in-place version (_compressCartesian) which modifies a and returns None.

Parameters

a ([zone, list of zones, base, pyTree]) – input data

Returns

identical to input

# - compressCartesian (pyTree) -
import Compressor.PyTree as Compressor
import Generator.PyTree as G
import Converter.Internal as Internal

a = G.cart((0,0,0), (1,1,1), (10,10,10))
Compressor._compressCartesian(a)
Internal.printTree(a)
Compressor.PyTree.uncompressCartesian(a)

Uncompress zones that has been compressed with compressCartesian. Exists also as an in-place version (_uncompressCartesian) which modifies a and returns None.

Parameters

a ([zone, list of zones, base, pyTree]) – input data

Returns

identical to input

# - uncompressCartesian (pyTree) -
import Compressor.PyTree as Compressor
import Generator.PyTree as G
import Converter.Internal as Internal

a = G.cart((0,0,0), (1,1,1), (10,10,10))
Compressor._compressCartesian(a)

Compressor._uncompressCartesian(a)
Internal.printTree(a)
Compressor.PyTree.compressCellN(a)

Compress cellN fields (valued 0,1,2).

Exists also as an in-place version (_compressCellN) which modifies a and returns None.

Parameters

a ([zone, list of zones, base, pyTree]) – input data

Returns

identical to input

# - compressCellN (pyTree) -
import Compressor.PyTree as Compressor
import Generator.PyTree as G
import Converter.PyTree as C
import Converter.Internal as Internal

a = G.cart((0,0,0), (1,1,1), (10,11,12))
C._initVars(a, '{centers:cellN}=1.')
Compressor._compressCellN(a)
Compressor._uncompressAll(a)
Internal.printTree(a)
C.convertPyTree2File(a, 'out.cgns')
Compressor.PyTree.compressCoords(a, tol=1.e-8, ctype=0)

Compress zone coordinates with sz library with a relative tolerance tol.

Exists also as an in-place version (_compressCoords) which modifies a and returns None.

Parameters

  • a ([zone, list of zones, base, pyTree]) – input data

  • tol (float) – control relative error on output

  • ctype (0 (sz), 1 (zfp)) – compression algorithm

Returns

identical to input

# - compressCoords (pyTree) -
import Compressor.PyTree as Compressor
import Generator.PyTree as G
import Converter.PyTree as C
import Converter.Internal as Internal

a = G.cart((0,0,0), (1,1,1), (5,3,4))
Compressor._compressCoords(a, tol=1.e-7, ctype=0)
Compressor._uncompressAll(a)
Internal.printTree(a)
C.convertPyTree2File(a, 'out.cgns')
Compressor.PyTree.compressFields(a, tol=1.e-8)

Compress zone fields with sz library with a relative tolerance tol.

Exists also as an in-place version (_compressFields) which modifies a and returns None.

Parameters

  • a ([zone, list of zones, base, pyTree]) – input data

  • tol (float) – control relative error on output

  • ctype (0 (sz), 1 (zfp)) – compression algorithm

Returns

identical to input

# - compressFields (pyTree) -
import Compressor.PyTree as Compressor
import Generator.PyTree as G
import Converter.PyTree as C
import Converter.Internal as Internal

a = G.cart((0,0,0), (1,1,1), (10,10,10))
C._initVars(a, '{F}={CoordinateX}')
C._initVars(a, '{centers:G}={centers:CoordinateY}')
Compressor._compressFields(a, tol=1.e-6, ctype=0)
Compressor._uncompressAll(a)
Internal.printTree(a)
C.convertPyTree2File(a, 'out.cgns')
Compressor.PyTree.compressElements(a)

Compress zone elements (connectivity).

Exists also as an in-place version (_compressElements) which modifies a and returns None.

Parameters

a ([zone, list of zones, base, pyTree]) – input data

Returns

identical to input

# - compressElements (pyTree) -
import Compressor.PyTree as Compressor
import Generator.PyTree as G
import Converter.PyTree as C
import Converter.Internal as Internal

a = G.cartHexa((0,0,0), (1,1,1), (25,23,24))
Compressor._compressElements(a)
Internal.printTree(a)
Compressor._uncompressAll(a)
Internal.printTree(a)
C.convertPyTree2File(a, 'out.cgns')
Compressor.PyTree.uncompressAll(a, tol=1.e-8)

Uncompress zones compressed with the previous compressors.

Exists also as an in-place version (_uncompressAll) which modifies a and returns None.

Parameters

a ([zone, list of zones, base, pyTree]) – input data

Returns

identical to input

# - uncompressAll (pyTree) -
import Compressor.PyTree as Compressor
import Generator.PyTree as G
import Converter.PyTree as C

a = G.cart((0,0,0), (1,1,1), (10,10,10))
C._initVars(a, '{F}={CoordinateX}')
C._initVars(a, '{centers:G}={centers:CoordinateY}')
Compressor._compressCoords(a, tol=1.e-6)
Compressor._compressFields(a, tol=1.e-6)
Compressor._uncompressAll(a)
C.convertPyTree2File(a, 'out.cgns')