Revisions to Different ways of modelling interactions between continuous and categorical predictors in GAM

replaced figure with one that is purely descriptive and not model-based.

edited Jul 19, 2021 at 12:08

1.3k
1
9
26

These are the mean color spectra at the locality level for the two species (rolling means were used):

For clarity, each line in the figure above represents the meanEach color spectrum predicted for each location withrefers to a separate GAM of form density~s(wl) based on samples of ~10 flowersdifferent species. The gray areas represent 95% CI for each GAMEach line refers to a different locality.

My final goal is to model the (potentially interactive) effect of Taxon and wavelength wl on % reflectance (referred to as density in the code and dataset) while accounting for Locality as a random effect in a mixed-effect GAM. For the moment I won't add the mixed effect part to my plate, which is already full enough with trying to understand how to model interactions.

For clarity, each line in the figure above represents the mean color spectrum predicted for each location with a separate GAM of form density~s(wl) based on samples of ~10 flowers. The gray areas represent 95% CI for each GAM.

My final goal is to model the (potentially interactive) effect of Taxon and wavelength wl on reflectance (referred to as density in the code and dataset) while accounting for Locality as a random effect in a mixed-effect GAM. For the moment I won't add the mixed effect part to my plate, which is already full enough with trying to understand how to model interactions.

These are the mean color spectra at the locality level for the two species (rolling means were used):

Each color refers to a different species. Each line refers to a different locality.

My final goal is to model the (potentially interactive) effect of Taxon and wavelength wl on % reflectance (referred to as density in the code and dataset) while accounting for Locality as a random effect in a mixed-effect GAM. For the moment I won't add the mixed effect part to my plate, which is already full enough with trying to understand how to model interactions.

Tweeted twitter.com/StackStats/status/1119254624819609600

occurred Apr 19, 2019 at 15:01

added 64 characters in body

Source Link

edited Apr 19, 2019 at 13:41

Marco Plebani

1.3k
1
9
26

The next model is the one that I have trouble understanding:

deleted 25 characters in body

Source Link

edited Apr 19, 2019 at 13:13

Marco Plebani

1.3k
1
9
26

My understanding is that theThe parametric part is the same for both species, but different splines are fitted for each species. It is a bit confusing to have a parametric part in the summary of GAMs, which are non-parametric. @IsabellaGhement explains:

Now, each species also have its own parametric estimate. To draw a risky comparison with linear models, my understanding is that if gam.interaction0 were a linear model it would be fitting different slopes but a common intercept for each species, while gam.interaction1 would allow for different intercepts and different slopes for each species.

My understanding is that the parametric part is the same for both species, but different splines are fitted for each species. It is a bit confusing to have a parametric part in the summary of GAMs, which are non-parametric. @IsabellaGhement explains:

Now, each species also have its own parametric estimate. To draw a risky comparison with linear models, my understanding is that if gam.interaction0 were a linear model it would be fitting different slopes but a common intercept for each species, while gam.interaction1 would allow for different intercepts and different slopes for each species.

The parametric part is the same for both species, but different splines are fitted for each species. It is a bit confusing to have a parametric part in the summary of GAMs, which are non-parametric. @IsabellaGhement explains:

Now, each species also have its own parametric estimate.