Rev Language Reference
mvUpDownSlide - Up-Down Proposal for Joint Sliding-Window Adjustments to Multiple Parameters
Simultaneously applies a sliding adjustment to multiple parameters, potentially
increasing some and decreasing others by the same amount.
Usage
Arguments
| delta : | RealPos (pass by value) |
| The scaling factor (strength) of the proposal. | |
| Default : 1 | |
| tune : | Bool (pass by value) |
| Should we tune the scaling factor during burnin? | |
| Default : TRUE | |
| weight : | RealPos (pass by value) |
| The weight determines the relative frequency with which this move will be attempted. For details, see the description of the 'moveschedule' parameter on the documentation page for 'mcmc()'. | |
| Default : 1 | |
| tuneTarget : | Probability (pass by value) |
| The acceptance probability targeted by auto-tuning. | |
| Default : 0.44 |
Details
This move adds a random value to a set of parameters, and optionally subtracts
the same value from another set of parameters. This may improve the mixing
of parameters that we expect to be either positively or negatively correlated,
such as speciation and extinction rates or the clock rate and branch durations
in time tree inference. The parameters to be incremented and decremented can be
selected using the `.addVariable()` and `.removeVariable()` methods. The actual
value to be added or subtracted is equal to delta * (u - 0.5), where delta is
a tuning parameter and u is a random draw from the uniform distribution on
[0, 1].
Example
moves = VectorMoves()
log_speciation_rate ~ dnNormal(-1, 0.5)
log_extinction_rate ~ dnNormal(-1, 0.5)
# Define the basic properties of the move
delta_up_down_move = mvUpDownSlide(delta=0.05, weight=5.0)
# Add variables to the move to account for their positive correlation
delta_up_down_move.addVariable(log_speciation_rate, up=TRUE)
delta_up_down_move.addVariable(log_extinction_rate, up=TRUE)
# Apply the move
moves.append( delta_up_down_move )
Edit this page