tsipkens / atems Goto Github PK

Description
The current PCM function has an output that differs subtly from the original implementation of Dastanpour et al.

To Reproduce
Direct reproduction is challenging, as the original code had different inputs and involved user intervention. Bug was found by inspection of pair correlation function (PCF) curves and the underlying code.

Expected behavior
The code should have the ability to reproduce previous version of the PCM method.

Other suggested changes
Consider adding versioning, to allow for current version, past version, and future versions, which may improve interpretation of the PCF curve. This can also be used to encode the simple and advanced versions of the existing PCM output, simplifying output to a single primary particle size depending on the specified version.

This example won't run as written, because the opts variable isn't yet defined.

If a user just copies the text and pastes into Matlab, it won't work.

When I run tools.write_json(Aggs, 'test.json');. Fails with

Error using jsonencode
Unable to encode objects of class logical as JSON-formatted text.
Error in tools.write_json (line 9)
t0 = jsonencode(var); % generate json text using built-in function

I think this is a Matlab version incompatibility. Here is my Matlab version info:

MATLAB Version: 9.14.0.2254940 (R2023a) Update 2
Operating System: macOS Version: 12.6.9 Build: 21G726
Java Version: Java 1.8.0_202-b08 with Oracle Corporation Java HotSpot(TM) 64-Bit Server VM mixed mode

Several of the main_* functions will occasionally lock up Matlab on my computer. main_kmeans and the final section of main_0 after agg.analyze_binary. I can step through each function line-by-line, so it's not clear where the problem is.

Here is my Matlab and operating system information:

MATLAB Version: 9.14.0.2254940 (R2023a) Update 2
Operating System: macOS Version: 12.6.9 Build: 21G726
Java Version: Java 1.8.0_202-b08 with Oracle Corporation Java HotSpot(TM) 64-Bit Server VM mixed mode

These lines refer to agg.seg_otsu_rb_orig(...) and agg.seg_otsu_rb(...). There aren't these functions in the agg directory. I think these should be agg.seg_otsu_orig(...) and agg.seg_otsu(...).

README refers to pp.kook*(...). I think this should be pp.hough_kook*.

This example code won't run as written, because fname is not yet defined. What is the structure of fname?

I was able to run this command:

Aggs = agg.analyze_binary(imgs_binary, pixsizes, imgs)

Based on the help in agg.analyze_binary, it looks like agg.analyze_binary(images_binary, Imgs), where Imgs is the structure output from tools.load_images should work, and should then include the file name, but I get an error:

Operator '*' is not supported for operands of type 'struct'.
Error in agg.analyze_binary (line 198)
Aggs0(jj).length = max((max(row)-min(row)), (max(col)-min(col))) * pixsize(ii);

• Change "more particles or add to existing particles" button to be labeled something like "Interactive tool to adjust thresholding"
  • Adjusting the "slider" does not seem to offer any feedback on how it is changing the threshold. That makes it very hard to judge whether the change is sufficient. I suggest overlying a color that shows the thresholded region.
  • I'm not sure if I'm doing it incorrectly, but every time I adjust the slider, I get a black image afterward, suggesting that the threshold is incorrect
  • There does not seem to be a way to go back in the steps. That would be helpful
  • Now I see that "Add (to) aggregates" button is sort of like a "start over" option. It is unclear what that actually does