Conversation
Reduces the parameter ranges for aberration coefficient and particle radii, removed optical axis position parameter from the Optuna search.
| " f\"Z: {particle.z()} vs {best_params['z']}\"\n", | ||
| " f\"{'Parameter':<15} {'Predicted':<20} {'Ground Truth':<20}\\n\"\n", | ||
| " f\"{'-' * 55}\\n\"\n", | ||
| " f\"{'Aberration':<15} {aberration.name():<20} {best_params['aberration_name']}\\n\"\n", |
There was a problem hiding this comment.
issue (blocking): The validation table currently labels the reference values (aberration and particle) as Predicted and the optimized values (best_params) as Ground Truth.
| @@ -114,7 +122,7 @@ | |||
| "\n", | |||
| "# Define optics.\n", | |||
| "optics = dt.Fluorescence(\n", | |||
| " magnification=10,\n", | |||
| " magnification=15,\n", | |||
There was a problem hiding this comment.
issue (non-blocking): the description states that the microscope has a pixel size of 0.1 um, but with resolution=1e-6 and magnification=15, the calculated pixel size is approximately 0.0667 um
|
I added some suggestions for updating the tutorial in a separate PR: #529 |
There was a problem hiding this comment.
Would it be better to rearrange the text description in sections 4 and 5? For example, RMSE is employed in the next section's code but is mentioned here in the text. or merger section 4 and 5 together.
4 participants
Reduces the parameter ranges for aberration coefficient and particle radii, removed optical axis position parameter from the Optuna search.