search

florianhartig / DHARMa

Diagnostics for HierArchical Regession Models
http://florianhartig.github.io/DHARMa/
200 stars 21 forks source link

Seeking Explanation for Residuals vs. Predicted Plot #382

Closed ToorSingh closed 1 year ago

ToorSingh commented 1 year ago

@florianhartig

Hi Florian,

I hope you're doing well. I recently conducted an analysis on a set of GLM models (binomial distribution with 0/1 responses) using DHARMa. The results of tests such as the KS test, dispersion test, and outlier test were all non-significant. However, when I plotted the residuals against model predictions, I encountered unexpected patterns. Interestingly, individual plots for each predictor variable showed no issues, and both zero inflation and spatial autocorrelation were also non-significant. Could you provide some insight into this matter? Please feel free to ask me if any other information is required.

Thank you, Tehlu

Model 1: image

Model 2:

image

Model 3:

image

florianhartig commented 1 year ago

Hello Toor,

by errors, you mean that there is a pattern in the res ~ pred plot?

What I can say is that your plots don't seem particularly concerning to me - yes there is a slightly wobbly pattern, but given the relatively low number of data points, this could also be random.

Best, F

ToorSingh commented 1 year ago

Yeah, the pattern! Thanks Florian.