Use results from the Bayesian interpretation of the GAM to obtain uncertainty estimates. See Wood (2006).

1 2 | ```
dsm.var.gam(dsm.obj, pred.data, off.set, seglen.varname = "Effort",
type.pred = "response")
``` |

`dsm.obj` |
an object returned from running |

`pred.data` |
either: a single prediction grid or list of prediction grids. Each grid should be a |

`off.set` |
a a vector or list of vectors with as many elements as there are in |

`seglen.varname` |
name for the column which holds the segment length (default value |

`type.pred` |
should the predictions be on the "response" or "link" scale? (default |

This is based on `dsm.var.prop`

taken from code by Mark Bravington and Sharon Hedley.

a list with elements

`model` | the fitted model object |

`pred.var` | variance of the regions given
in `pred.data` . |

`bootstrap` | logical, always `FALSE` |

`model` | the fitted model with the extra term |

`dsm.object` | the original model, as above |

David L. Miller

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | ```
## Not run:
library(Distance)
library(dsm)
# load the Gulf of Mexico dolphin data (see ?mexdolphins)
data(mexdolphins)
attach(mexdolphins)
# fit a detection function and look at the summary
hr.model <- ds(distdata, max(distdata$distance),
key = "hr", adjustment = NULL)
summary(hr.model)
# fit a simple smooth of x and y
mod1 <- dsm(N~s(x,y), hr.model, segdata, obsdata)
# Calculate the variance
mod1.var <- dsm.var.gam(mod1, preddata, off.set=preddata$area)
# this will give a summary over the whole area in mexdolphins$preddata
# detach the data
detach("mexdolphins")
## End(Not run)
``` |

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.