Changes for next time

[ChasNelson1990]

Hi all, please add comments for everything we'd like to change for next time.

[ChasNelson1990]

@ChasNelson1990 needs less content/more time for the image processing in Python. Probably worth removing deconvolution entirely and having it in another session. Also worth emphasising the helical nature of tasks, e.g. use 'Tasks' and 'Exercises' and 'Challenges' like in prerequisite course.

[ChasNelson1990]

Swap @todd-fallesen's session on Fiji to Python with @ChasNelson1990's session on Image Processing

[ChasNelson1990]

Please could everybody use full variable names, e.g. myParticleImage not jut img. It's better for people reading outside of the course.

[ChasNelson1990]

@dwaithe to update first image in images-in-python notebook (number alignment)

[ChasNelson1990]

Please could everybody use full variable names, e.g. myParticleImage not jut img. It's better for people reading outside of the course.

Or snake case my_particle_image. I just mean informative names, camelCase or snake_case is a personal preference in my mind (let's ignore PEP-8)

[ChasNelson1990]

Some of @todd-fallesen's course is covered in the prerequisites. I think it's great to repeat and revise this but it's worth bearing this in mind if we want to cut time somewhere.

[ChasNelson1990]

I think @ChasNelson1990's session on image processing should perhaps follow a more pipeline orientated approach like @todd-fallesen's.

[ChasNelson1990]

@todd-fallesen finished 30 minutes early (well, started bio-relevant section 30 minutes before the end). I suggest taking that time away and giving it to @ChasNelson1990's image-processing section but polishing the biologically relevant example as a resource for post-course revision?

[ChasNelson1990]

Labels on @dwaithe's slides are very small

[gchennell]

overlay local maxima on raw image as well for clarity

…

On Tue, 10 Dec 2019 at 10:19, Chas Nelson ***@***.***> wrote: Labels on @dwaithe <https://github.com/dwaithe>'s slides are very small — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#5>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB46Y5TJTUS7YQ325CIYNP3QX5ULJANCNFSM4JYLHSGA> .

-- Sent from my PC

[ChasNelson1990]

Turn @ChasNelson1990's processing session specifically into 'preprocessing'?

[ChasNelson1990]

All to use regionprops_table and pandas.DataFrame to make visualising, describing, filtering, etc. the results more simple and more visual (requires conda to update their scikit-image).

[ChasNelson1990]

Should @dwaithe's file formats be on the first day? Along with Python cmaps?