Carlin Liao
Carlin Liao
See [these](https://github.com/histocartography/patho-quant-explainer/issues/9) [three](https://github.com/histocartography/patho-quant-explainer/issues/8) [issues](https://github.com/histocartography/patho-quant-explainer/issues/7) in the patho-quant-explainer repository. To summarize, I was trying to reproduce the results shown in the pathology quantitative explainer paper, but failed to do so. After...
When using the GraphGradCAMExplainer, we use an pretrained torch GNN model set to eval mode since we're no longer training the model. However, to find the node importances, the Explainer...
For some math statements, the line numbering appears to count an overlapping ghost line somewhere around the end of the align/gather/alignat statement. ``` \begin{linenomath} \begin{alignat*}{3} &\min_{x,h} \qquad && \sum_{(i,j) \in...
Consider the following code block from `generate_cell_graphs.py`: ``` for image_path in tqdm(image_fnames): # ... excised for brevity # if file was not already created + not too big + not...
I first noticed this issue in `run_qualitative_analysis.py` where it would error in main at line 315 ``` concepts = [cg.ndata['concepts'].cpu().detach().numpy() for cg in cell_graphs] ``` because a cell graph wouldn't...
I've been trying to replicate the patho quant results using the BRACS data (i.e., no changes to the source code or function calls), but have been failing at multiple points....
We also discussed how the `[upload-importance]` section in a graph config file should be generated instead of manually written by the user, at least in Nextflow workflow, since the plugin...
Current default behavior is to use the `latest` image tag for the SPT docker image in Singularity and Docker Nextflow workflows, but we observed an issue where Singularity doesn't check...
Since we're adapting pathology datasets for use in machine learning, often our datasets end up being imbalanced, e.g., the treatment non-responding set ends up having 2x the number of graphs...
I often find myself editing the `workflow` Makefiles when I'm investigating issues with `spt workflow configure` and `nextflow run` so I can see what's wrong with the processes being run...