Objects
Image
Modified instances of multiscaleimage.MultiscaleImage.
Modification: Each scale has at least one named affine transform parameter attached to it as further data variables next to scale<scale>/image.
While instances of multiscale-spatial-image can be serialized to and from NGFF, modified instances of multiscaleimage.MultiscaleImage as used by multiview-stitcher cannot (yet) be serialized to and from NGFF (see here), as the support for affine transforms is missing.
Example string representation:
print(msims[0])
DataTree('None', parent=None)
│ Dimensions: ()
│ Data variables:
│ *empty*
│ Attributes:
│ multiscaleSpatialImageVersion: 1
│ multiscales: [{'@type': 'ngff:Image', 'axes': [{'name'...
├── DataTree('scale0')
│ Dimensions: (t: 1, x_in: 4, x_out: 4, c: 1, z: 179, y: 1040, x: 1392)
│ Coordinates:
│ * c (c) int64 0
│ * t (t) int64 0
│ * x (x) float64 0.0 0.645 1.29 1.935 ... 895.9 896.6 897.2
│ * y (y) float64 0.0 0.645 1.29 1.935 ... 668.9 669.5 670.2
│ * z (z) float64 0.0 2.58 5.16 7.74 ... 451.5 454.1 456.7 459.2
│ Dimensions without coordinates: x_in, x_out
│ Data variables:
│ affine_metadata (t, x_in, x_out) float64 1.0 0.0 0.0 0.0 ... 0.0 0.0 1.0
│ image (t, c, z, y, x) uint16 dask.array<chunksize=(1, 1, 179, 256, 256), meta=np.ndarray>
├── DataTree('scale1')
│ Dimensions: (t: 1, x_in: 4, x_out: 4, c: 1, z: 179, y: 520, x: 696)
│ Coordinates:
│ * c (c) int64 0
│ * t (t) int64 0
│ * x (x) float64 0.3225 1.613 2.902 4.193 ... 894.3 895.6 896.9
│ * y (y) float64 0.3225 1.613 2.902 4.193 ... 667.3 668.5 669.8
│ * z (z) float64 0.0 2.58 5.16 7.74 ... 451.5 454.1 456.7 459.2
│ Dimensions without coordinates: x_in, x_out
│ Data variables:
│ affine_metadata (t, x_in, x_out) float64 1.0 0.0 0.0 0.0 ... 0.0 0.0 1.0
│ image (t, c, z, y, x) uint16 dask.array<chunksize=(1, 1, 179, 256, 256), meta=np.ndarray>
└── DataTree('scale2')
Dimensions: (t: 1, x_in: 4, x_out: 4, c: 1, z: 179, y: 260, x: 348)
Coordinates:
* c (c) int64 0
* t (t) int64 0
* x (x) float64 0.9675 3.548 6.128 8.707 ... 891.1 893.6 896.2
* y (y) float64 0.9675 3.548 6.128 8.707 ... 664.0 666.6 669.2
* z (z) float64 0.0 2.58 5.16 7.74 ... 451.5 454.1 456.7 459.2
Dimensions without coordinates: x_in, x_out
Data variables:
affine_metadata (t, x_in, x_out) float64 1.0 0.0 0.0 0.0 ... 0.0 0.0 1.0
image (t, c, z, y, x) uint16 dask.array<chunksize=(1, 1, 179, 256, 256), meta=np.ndarray>
Transformation parameters
Affine transformation parameters
xarray.DataArray containing parameters in the form of
- a homogeneous transform matrix
- of dimensionality (ndim+1, ndim+1)
- datatype float
with axis labels - 't' - 'x_in' - 'x_out'
Example string representation:
print(msims[0]['scale0/affine_manual'])