Convert line and fill types

Convert contour lines to a different line type

convert_lines() is used to convert contour lines to a different LineType.

The following example creates two contour lines in LineType.Separate format where each line’s 2D points are in separate NumPy arrays:

>>> from contourpy import contour_generator, convert_lines
>>> cont_gen = contour_generator(z=[[0, 1, 0], [0, 1, 0]], line_type="Separate")
>>> lines = cont_gen.lines(0.5)
>>> lines
[array([[0.5, 1.], [0.5, 0.]]), array([[1.5, 0.], [1.5, 1.]])]

This can be converted to LineType.ChunkCombinedOffset:

>>> lines = convert_lines(lines, cont_gen.line_type, "ChunkCombinedOffset")
>>> lines
([array([[0.5, 1.], [0.5, 0.], [1.5, 0.], [1.5, 1.]])], [array([0, 2, 4], dtype=uint32)])

in which the lines are combined into a single array of 2D points and the offset array shows the start and end offsets of each line.

Any LineType can be converted to any other LineType. When converting from a non-chunked line type (LineType.Separate or LineType.SeparateCode) to a chunked one (LineType.ChunkCombinedCode, LineType.ChunkCombinedOffset or LineType.ChunkCombinedNan), all lines are placed together in the first chunk. When converting in the other direction, all chunk information is discarded as all lines are appended to the same list (LineType.Separate) or lists (LineType.SeparateCode).

Convert filled contours to a different fill type

convert_filled() is used to convert filled contours to a different FillType.

The following example creates two filled polygons in FillType.OuterCode format where each of the polygons points and codes are in separate NumPy arrays:

>>> from contourpy import contour_generator, convert_filled
>>> cont_gen = contour_generator(z=[[0, 1, 0], [0, 1, 0]], fill_type="OuterCode")
>>> filled = cont_gen.filled(0.3, 0.6)
>>> filled
([array([[0.3, 0.], [0.6, 0.], [0.6, 1.], [0.3, 1.], [0.3, 0.]]),
  array([[1.4, 0.], [1.7, 0.], [1.7, 1.], [1.4, 1.], [1.4, 0.]])],
 [array([1, 2, 2, 2, 79], dtype=uint8),
  array([1, 2, 2, 2, 79], dtype=uint8)])

This can be converted to FillType.ChunkCombinedOffsetOffset:

>>> filled = convert_filled(filled, cont_gen.fill_type, "ChunkCombinedOffsetOffset")
>>> filled
([array([[0.3, 0.], [0.6, 0.], [0.6, 1.], [0.3, 1.], [0.3, 0.],
         [1.4, 0.], [1.7, 0.], [1.7, 1.], [1.4, 1.], [1.4, 0.]])],
 [array([0, 5, 10], dtype=uint32)],
 [array([0, 1, 2], dtype=uint32)])

in which the points are combined into a single 2D array and there are two arrays of offsets, one for the boundary offsets and one for the polygon (outer boundary) offsets.

Not all FillType can be converted to any other FillType. Two of the FillType (FillType.ChunkCombinedCode and FillType.ChunkCombinedOffset) do not include information about the relationship between outer and inner boundaries so they cannot be converted to any of the FillType that need this information (FillType.OuterCode, FillType.OuterOffset, FillType.ChunkCombinedCodeOffset and FillType.ChunkCombinedOffsetOffset).

Also, when converting from a non-chunked fill type (FillType.OuterCode or FillType.OuterOffset) to a chunked one (any of the others), all polygons are placed together in the first chunk. When converting in the other direction, all chunk information is discarded as all polygons are appended to the same lists.