Skip to contents

(Vignette under construction!)

Mapping from component inputs and latent states to component effects

main, group, replicate

Yes

No

input

mapper

weights

combined inputs:
list(mapper = list(main, group, replicate),
scale = weights)

mapper_multi
list(main, group, replicate)

mapper_marginal

mapper_scale

mapper_pipe

state

mapper_linearised
(offset, jacobian, state0)

effect

linearise

ibm_linear

ibm_eval

ibm_eval

state0

Linearising a mapping

mapper

input

state0

offset

jacobian

mapper_linearised
(offset, jacobian, state0)

ibm_amatrix

ibm_linear

ibm_jacobian

ibm_eval

Component input evaluation

For each <label> of main, group, replicate, and weights, the given expression expr is evaluated in the data context, producing the input to the component mapper. For spatial covariate inputs, the corresponding <label>_layer expression is also evaluated.

Red nodes indicate deprecated behaviour retained for backwards compatibility.

error

TRUE

FALSE

function

Yes

No

vector,
matrix,
data.frame
list

formula

spatial
covariate

label_expr

data

.envir

label_input

input = NULL

input = 1

eval

value=fun(.data.)

model.Matrix

eval_spatial

SpatialPoints(value, crs)

crs=fm_CRS(.data.)

Type of
result

null_on_fail?

Is expr =
coordinates?

label_layer_expr

label_layer

eval

Intergration point construction

Flow diagram for new integration scheme construction, implemented as fm_int(domain, samplers) methods.

multi sampler

single sampler

for each

for each

for each

samplers

domains

multi domain
samplers

single domain
samplers

remove sampler
domains

full domain
samplers

compute ips
for each row

ips

compute ips
for each domain

ips
for each domain

cprod within
each sampler row

ips

ips list

cprod

joint ips