Batch analysis: tapestries - galizia-lab/pyview GitHub Wiki
Overviews, when there are many, are called tapestries.
In VIEW and ILTIS, we have looked at the traces. Stimulus is at frame 25, peak response in most ROIs is around frame 45. Let’s calculate the difference between 45 and 25, using CTV 22 (as experimented in VIEW, see above).
Similar to the situation for movies, do the following:
In the folder “06_PROGS”, you will find the file create_tapestry_synthetic.py.
(For Windows: make sure you use the latest version of create_tapestry_synthetic.py')
You can run this file directly in python (command line, or calling it from a python editor), within the right environment.
You should adapt this file to your own data. Most importantly, for each animal, you have to create a layout (the order of images in the
tapestry). And you need to specify flags, in particular for the limits of false-color coding. Therefore, each animal needs its own tapestry_animal.yml file. In the file create_tapestry_synthetic.py, you will then list all tapestry .yml files
- Specify which animals/.yml combinations are to be evaluated (here it is “06_PROGS/tapestry_ctv22_individual.yml”, and a second tapestry “06_PROGS/tapestry_ctv22_global.yml”)
- You can add options for additional text to be written next to the overviews.
Run the python program – it will create a folder “04_OUTPUT/tapestries/”. In that folder, we copied all relevant files (the python program and the .yml files used, for later reference), a .html page for each tapestry, and subfolders for the individual images, so that they can be used for images in publications.
Spatial filtering is controlled by Signal_FilterSpaceFlag and Signal_FilterSpaceSize (in our example: value 1).
Example 1
In our example: tapestry_ctv22_individual (flag: SO_individualScale: 3) shows overviews where every overview is scaled individually. You will see consistent response patterns for POS+2 to POS+20, increasing concentration does not change the spatial pattern too much, though at lower concentrations the effect of noise is more visible. There is a negative pattern for POS-10, but all the (fictive) concentrations around 0 (NOISE) are so dominated by noise that the response pattern is not visible. Next to the individual overviews there are the scaling parameters, and at the bottom the maximum and minimum values across all overviews. Use these for deciding limits in the next tapestry.
Example 2
In our example, tapestry_ctv22_global (flag: SO_individualScale: 0) has limits of -0.2 and 0.2 (in the “06_PROGS/tapestry_ctv22_global.yml” file). In the tapestry the increasing responses with increasing concentrations are clearly visible.
| Previous page | Contents | Next page |
|---|---|---|
| Batch analysis: movies | Contents | Batch analysis: traces |