Explanation of Positional Variance Diagrams and Color Transparency

[gy1818]

I’m working with positional variance diagrams and need help understanding the meaning of different colors with varying levels of transparency. Here’s some context:

I defined Z_vals = np.array([[1, 2, 3]] * N).

I have 26 Regions of Interest (ROIs).

My prob_ml_stage is very low, not even reaching 0.5.

In the diagrams, the colors (red, pink, purple) vary in transparency. Can someone explain what these different levels of transparency represent, especially in relation to my data?

[gy1818]

[katrinaCode]

Hi @gy1818, not a sustain dev but as a quick/simple answer for interpretation: you can imagine the PVDs like viewing sets of probability distributions from a bird's eye view. For simplicity, I imagine them as normal distributions/Gaussians of probabilities centred around stages, but that isn't necessarily mathematically correct. However, it is sufficient as a basic explainer.

The more narrow the peak of the Gaussian (reflecting a greater probability that a certain event happens at a certain stage), the more opaque the colour; the broader the Gaussian (reflecting a lesser probability that a certain event happens at a certain stage), the more transparent the colour. So in short, the intensity/opaqueness of the colour directly represents the probability of a certain event happening at a certain stage. You can imagine the "length" of a colour on the PVD to be analogous to the length of the tails of the Gaussian.

To improve your transparency issue (which is really a probability issue), a simple thing to do would be to reduce your number of stages. To reduce stages (whose number is equal to num biomarkers x num Z_vals, so 26 x 3 = 78 in your case), you could either reduce your number of biomarkers, or reduce the number of Z_vals.

For the meaning of the different colours, see: #49