Getting Started¶

Installation¶

You can easily install the package with pypi:

    pip install h5tiff

or you can install it with:

    pip install -e path_to_h5tiff/h5tiff

which allows you to customize the converter to your needs (see an example here).

Using and Customizing¶

This package was specifically written to be customized in order to process correctly differente h5 file strurctures:

At the moment there’re only two modes:

default
LFM

⚠ WARNING! The data used as input is supposed to be a list of images/stack which will be separated in single Tiff files.

Default mode¶

This is a very crude mode that tries to accept various cases underneath it.

The h5 file is opened expeting to find the data in a dictionary format, where the ‘Data’ key is where the data is stored.

    yourdata = h5file['Data'] 

Any other key is instead considered as metadata and inserted in a single dictionary. When the pictures are saved as Tiff files the metadata is added to each one.

LFM mode¶

This a specific mode which expects the data in the following format:

    yourdata = {
                'Data': list_of_images,
                'motorData' = motorData,
                'metadata' = metadata,
                'img_time' = img_time,
                }

where motorData is a list of information regarding the position of the stage motors of the light field microscope. When the tiff file is saved only the correct position is attached as metadata.
metadata is general information and specifics about the whole set of images. This information is attached to every tiff file.
Finally, img_time is a 2D list of time stamps (one before the picture was taken, one after).
Only the times corresponding to a specific image are attached in the metadata of that image.

Adding your own mode¶

To add a different mode to the code there are four modifications to be made:

You need to add the name of your mode here:

MODES = ("LFM", "DEFAULT", "ADD_HERE_YOUR_MODE_NAME")

Then, you need to write a specific function to call for loading and converting the images. You can add it at the end and use some of the previous code as a reference.
- Load function:
  - The funtion to load does’t need any input beause the path to the h5 file is requested when creating the converter object.
  - It has to return the data (list of images) and metadata (dictionary)
- Convert function:
  - It doesn’t need any inputs, and it returns nothing
  - We suggest to make use of the save_tiff function, which requires an image, a metadata dictionary, a saving name, and a saving path.
Finally, you need to connect the functions you wrote with the main processing pipeline. Just copy the examples that you find commented here (load), and here (convert)