<a name="qttorch.dok"/>
# QTTorch Package Reference Manual #

Package `qttorch` declares two functions
that convert `torch.Tensor` to `QImage` and vice-versa.

Loading this package automatically loads
packages [torch](..:torch:index) 
and [qt](..:qt:index).


<a name="qimagefromtensor"/>
### qt.QImage.fromTensor(tensor) ###

`qt.QImage.fromTensor(tensor)`

Return a new [QImage](..:qtgui:index#qimage) filled 
with data from a torch [Tensor](..:torch:index#Tensor).

Tensor `tensor` must have 2 or 3 dimensions.
  * The first dimension defines the image width.
  * The second dimension defines the image height.
  * The third dimension size must be 1, 3, or 4
for monochrome, rgb, and rgba images respectively.
Each component is a floating point number in range 0 to 1.


<a name="qimagetotensor"/>
### qimage:toTensor(arg) ###

`qimage.toTensor(tensor)`

Fill the [Tensor](..:torch:index#Tensor) `tensor` with data 
from [QImage](..:qtgui:index#qimage) `qimage`.
Tensor `tensor` must have 2 or 3 dimensions.
The first two dimensions must be equal to the 
image width and height respectively.
The third dimension must be 1, 3, or 4
for monochromatic, rgb or rgba images.
The `qimage` data will be converted to the desired format
and stored into the tensor, each component being
represented by a floating point number in range 0 to 1.
This function returns the filled tensor.

`qimage.toTensor(depth)`

Create a tensor with data from image `qimage`.
Argument `depth` must be 1, 3, or 4
for extracting monochromatic, rgb or rgba data.
When `depth` is one, this function returns a two-dimensional tensor.
Otherwise it returns a three-dimensional tensor with
the third dimension equal to `depth`. 
The first two dimensions are always 
the image width and height.