Struct sampling::wang_landau::WangLandau1T
source · pub struct WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy> { /* private fields */ }
Expand description
The 1/t Wang Landau approach comes from this paper
R. E. Belardinelli and V. D. Pereyra, “Fast algorithm to calculate density of states,” Phys. Rev. E 75: 046701 (2007), DOI 10.1103/PhysRevE.75.046701
- The original Wang Landau algorithm comes from this paper
F. Wang and D. P. Landau, “Efficient, multiple-range random walk algorithm to calculate the density of states,” Phys. Rev. Lett. 86, 2050–2053 (2001), DOI 10.1103/PhysRevLett.86.2050
Implementations§
source§impl<Hist, Rng, Ensemble, S, Res, Energy> WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
impl<Hist, Rng, Ensemble, S, Res, Energy> WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
sourcepub fn into_inner(self) -> (Ensemble, Hist, Rng)
pub fn into_inner(self) -> (Ensemble, Hist, Rng)
Returns internal ensemble, histogram and Rng
source§impl<Hist, R, E, S, Res, Energy> WangLandau1T<Hist, R, E, S, Res, Energy>where
Hist: Histogram + HistogramVal<Energy>,
impl<Hist, R, E, S, Res, Energy> WangLandau1T<Hist, R, E, S, Res, Energy>where
Hist: Histogram + HistogramVal<Energy>,
sourcepub fn is_initialized(&self) -> bool
pub fn is_initialized(&self) -> bool
Check if self
is initialized
- if this returns true, you can begin the WangLandau simulation
- otherwise call one of the
self.init*
methods
source§impl<Hist, R, E, S, Res, Energy> WangLandau1T<Hist, R, E, S, Res, Energy>
impl<Hist, R, E, S, Res, Energy> WangLandau1T<Hist, R, E, S, Res, Energy>
sourcepub fn set_initial_probability_guess(
self,
new_guess: Vec<f64>,
new_log_f: f64
) -> Result<Self, SetInitialError>where
Hist: Histogram,
pub fn set_initial_probability_guess(
self,
new_guess: Vec<f64>,
new_log_f: f64
) -> Result<Self, SetInitialError>where
Hist: Histogram,
Set the initial guess for the non-normalized probability estimate
new_guess
your new guess for the probability estimate. Its length has to equal the number of bins of the internal histogram which is the same as the length of the old estimate which you can get by calling log_density. All contained values have to be finitenew_log_f
: Which log_f to start at? 0.0 < log_f <= 10.0 has to be true. If you don’t know what’s best I recommand starting with log_f=1.0, the better your probability estimate is, the smaller this value can be
Note
This will reset the calculation. Meaning you will have to call one of the initializing functions like init_greedy_heuristic
again
and all internal counters are reset to 0
source§impl<Hist, R, E, S, Res, Energy> WangLandau1T<Hist, R, E, S, Res, Energy>
impl<Hist, R, E, S, Res, Energy> WangLandau1T<Hist, R, E, S, Res, Energy>
sourcepub fn new(
log_f_threshold: f64,
ensemble: E,
rng: R,
step_size: usize,
histogram: Hist,
check_refine_every: usize
) -> Result<Self, WangLandauErrors>
pub fn new( log_f_threshold: f64, ensemble: E, rng: R, step_size: usize, histogram: Hist, check_refine_every: usize ) -> Result<Self, WangLandauErrors>
Create a new WangLandau simulation
IMPORTANT You have to call one of the init*
functions,
to create a valid state, before you can start the simulation
Parameter
log_f_threshold
: how small should the ln(f) (see paper) become until the simulation is finished?ensemble
: The ensemble to explore. Current state of ensemble will be used as inital condition for theinit*
functionsstep_size
: The markov steps will be performed with this step size, e.g.,ensemble.m_steps(step_size)
histogram
: Provides the binning. You can either use one of the already implemented histograms, likeHistU32Fast
,HistU32
,HistF64
etc. or implement your own by implementing the traitsHistogram + HistogramVal<Energy>
yourselfcheck_refine_every
: how often to check, if every bin in the histogram was hit. Needs to be at least 1. Good values depend on the problem at hand, but if you are unsure, you can start with a value like 1000
sourcepub fn init_greedy_heuristic<F>(
&mut self,
energy_fn: F,
step_limit: Option<u64>
) -> Result<(), WangLandauErrors>
pub fn init_greedy_heuristic<F>( &mut self, energy_fn: F, step_limit: Option<u64> ) -> Result<(), WangLandauErrors>
Find a valid starting Point
- if the ensemble is already at a valid starting point, the ensemble is left unchanged (as long as your energy calculation does not change the ensemble)
- Uses a greedy heuristik. Performs markov steps. If that brought us closer to the target interval, the step is accepted. Otherwise it is rejected
Parameter
step_limit
: Some(val) -> val is max number of steps tried, if no valid state is found, it will return an Error. None -> will loop until either a valid state is found or foreverenergy_fn
function calculatingSome(energy)
of the system or rather the Parameter of which you wish to obtain the probability distribution. Has to be the same function as used for the wang landau simulation later. If there are any states, for which the calculation is invalid,None
should be returned- steps resulting in ensembles for which
energy_fn(&mut ensemble)
isNone
will always be rejected
sourcepub fn init_interval_heuristik<F>(
&mut self,
overlap: NonZeroUsize,
energy_fn: F,
step_limit: Option<u64>
) -> Result<(), WangLandauErrors>
pub fn init_interval_heuristik<F>( &mut self, overlap: NonZeroUsize, energy_fn: F, step_limit: Option<u64> ) -> Result<(), WangLandauErrors>
Find a valid starting Point
- if the ensemble is already at a valid starting point, the ensemble is left unchanged (as long as your energy calculation does not change the ensemble)
- Uses overlapping intervals. Accepts a step, if the resulting ensemble is in the same interval as before, or it is in an interval closer to the target interval
- Take a look at the
HistogramIntervalDistance
trait
Parameter
step_limit
: Some(val) -> val is max number of steps tried, if no valid state is found, it will return an Error. None -> will loop until either a valid state is found or foreverenergy_fn
function calculatingSome(energy)
of the system or rather the Parameter of which you wish to obtain the probability distribution. Has to be the same function as used for the wang landau simulation later. If there are any states, for which the calculation is invalid,None
should be returned- steps resulting in ensembles for which
energy_fn(&mut ensemble)
isNone
will always be rejected
sourcepub fn init_mixed_heuristik<F, U>(
&mut self,
overlap: NonZeroUsize,
mid: U,
energy_fn: F,
step_limit: Option<u64>
) -> Result<(), WangLandauErrors>where
F: Fn(&mut E) -> Option<Energy>,
Hist: HistogramIntervalDistance<Energy>,
U: One + Bounded + WrappingAdd + Eq + PartialOrd,
pub fn init_mixed_heuristik<F, U>(
&mut self,
overlap: NonZeroUsize,
mid: U,
energy_fn: F,
step_limit: Option<u64>
) -> Result<(), WangLandauErrors>where
F: Fn(&mut E) -> Option<Energy>,
Hist: HistogramIntervalDistance<Energy>,
U: One + Bounded + WrappingAdd + Eq + PartialOrd,
Find a valid starting Point
- if the ensemble is already at a valid starting point, the ensemble is left unchanged (as long as your energy calculation does not change the ensemble)
overlap
- seeHistogramIntervalDistance
trait Should be greater than 0 and smaller than the number of bins in your histogram. E.g.overlap = 3
if you have 200 binsmid
- should be something like128u8
,0i8
or0i16
. It is very unlikely that using a type with more than 16 bit makes sense for midstep_limit
: Some(val) -> val is max number of steps tried, if no valid state is found, it will return an Error. None -> will loop until either a valid state is found or forever- alternates between greedy and interval heuristik everytime a wrapping counter passes
mid
orU::min_value()
- I recommend using this heuristik, if you do not know which one to use
Parameter
energy_fn
function calculatingSome(energy)
of the system or rather the Parameter of which you wish to obtain the probability distribution. Has to be the same function as used for the wang landau simulation later. If there are any states, for which the calculation is invalid,None
should be returned- steps resulting in ensembles for which
energy_fn(&mut ensemble)
isNone
will always be rejected
sourcepub fn wang_landau_step<F>(&mut self, energy_fn: F)
pub fn wang_landau_step<F>(&mut self, energy_fn: F)
Wang Landau Step
- performs a single Wang Landau step
Parameter
energy_fn
function calculatingSome(energy)
of the system or rather the Parameter of which you wish to obtain the probability distribution. If there are any states, for which the calculation is invalid,None
should be returned- steps resulting in ensembles for which
energy_fn(&mut ensemble)
isNone
will always be rejected
Important
- You have to call one of the
self.init*
functions before calling this one - will panic otherwise
sourcepub unsafe fn wang_landau_step_unsafe<F>(&mut self, energy_fn: F)
pub unsafe fn wang_landau_step_unsafe<F>(&mut self, energy_fn: F)
Wang Landau Step
- if possible, use
self.wang_landau_step()
instead - it is safer - performs a single Wang Landau step
Parameter
energy_fn
function calculatingSome(energy)
of the system or rather the Parameter of which you wish to obtain the probability distribution. If there are any states, for which the calculation is invalid,None
should be returned- steps resulting in ensembles for which
energy_fn(&mut ensemble)
isNone
will always be rejected
Safety
- You have to call one of the
self.init*
functions before calling this one - will panic otherwise - unsafe, because you have to make sure, that the
energy_fn
function does not change the state of the ensemble in such a way, that the result ofenergy_fn
changes when called again. Maybe do cleanup at the beginning of the energy function?
sourcepub fn wang_landau_step_acc<F>(&mut self, energy_fn: F)
pub fn wang_landau_step_acc<F>(&mut self, energy_fn: F)
Wang Landau Step
- performs a single Wang Landau step
Parameter
energy_fn
function calculating the energy of the system on the fly- steps resulting in invalid ensembles are not allowed!
Important
- You have to call one of the
self.init*
functions before calling this one - will panic otherwise
sourcepub fn wang_landau_convergence<F>(&mut self, energy_fn: F)
pub fn wang_landau_convergence<F>(&mut self, energy_fn: F)
Wang Landau
- perform Wang Landau simulation
- calls
self.wang_landau_step(energy_fn, valid_ensemble)
untilself.is_finished()
sourcepub fn wang_landau_convergence_acc<F>(&mut self, energy_fn: F)
pub fn wang_landau_convergence_acc<F>(&mut self, energy_fn: F)
Wang Landau - efficient energy calculation
- perform Wang Landau simulation
- calls
self.wang_landau_step_acc(energy_fn, valid_ensemble)
untilself.is_finished()
sourcepub unsafe fn wang_landau_convergence_unsafe<F>(&mut self, energy_fn: F)
pub unsafe fn wang_landau_convergence_unsafe<F>(&mut self, energy_fn: F)
Wang Landau
- if possible, use
self.wang_landau_convergence()
instead - it is safer - perform Wang Landau simulation
- calls
self.wang_landau_step_unsafe(energy_fn, valid_ensemble)
untilself.is_finished()
Safety
- You have mutable access to your ensemble, which is why this function is unsafe. If you do anything, which changes the future outcome of the energy function, the results will be wrong! I use the unsafe keyword here to force the user to acknowledge that.
sourcepub fn wang_landau_while<F, W>(&mut self, energy_fn: F, condition: W)
pub fn wang_landau_while<F, W>(&mut self, energy_fn: F, condition: W)
Wang Landau
- perform Wang Landau simulation
- calls
self.wang_landau_step(energy_fn)
untilself.is_finished()
orcondition(&self)
is false
sourcepub fn wang_landau_while_acc<F, W>(&mut self, energy_fn: F, condition: W)
pub fn wang_landau_while_acc<F, W>(&mut self, energy_fn: F, condition: W)
Wang Landau
- perform Wang Landau simulation
- calls
self.wang_landau_step(energy_fn)
untilself.is_finished()
orcondition(&self)
is false
sourcepub unsafe fn wang_landau_while_unsafe<F, W>(
&mut self,
energy_fn: F,
condition: W
)
pub unsafe fn wang_landau_while_unsafe<F, W>( &mut self, energy_fn: F, condition: W )
Wang Landau
- if possible, use
self.wang_landau_while()
instead - it is safer - perform Wang Landau simulation
- calls
self.wang_landau_step(energy_fn)
untilself.is_finished()
orcondition(&self)
is false
Safety
- You have mutable access to your ensemble, which is why this function is unsafe. If you do anything, which changes the future outcome of the energy function, the results will be wrong! I use the unsafe keyword here to force the user to acknowledge that
Trait Implementations§
source§impl<Hist: Clone, Rng: Clone, Ensemble: Clone, S: Clone, Res: Clone, Energy: Clone> Clone for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
impl<Hist: Clone, Rng: Clone, Ensemble: Clone, S: Clone, Res: Clone, Energy: Clone> Clone for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
source§fn clone(&self) -> WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
fn clone(&self) -> WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
Returns a copy of the value. Read more
1.0.0 · source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
Performs copy-assignment from
source
. Read moresource§impl<Hist: Debug, Rng: Debug, Ensemble: Debug, S: Debug, Res: Debug, Energy: Debug> Debug for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
impl<Hist: Debug, Rng: Debug, Ensemble: Debug, S: Debug, Res: Debug, Energy: Debug> Debug for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
source§impl<'de, Hist, Rng, Ensemble, S, Res, Energy> Deserialize<'de> for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>where
Hist: Deserialize<'de>,
Rng: Deserialize<'de>,
Ensemble: Deserialize<'de>,
S: Deserialize<'de>,
Energy: Deserialize<'de>,
impl<'de, Hist, Rng, Ensemble, S, Res, Energy> Deserialize<'de> for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>where
Hist: Deserialize<'de>,
Rng: Deserialize<'de>,
Ensemble: Deserialize<'de>,
S: Deserialize<'de>,
Energy: Deserialize<'de>,
source§fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
Deserialize this value from the given Serde deserializer. Read more
source§impl<Hist, R, E, S, Res, Energy> GlueAble<Hist> for WangLandau1T<Hist, R, E, S, Res, Energy>where
Hist: Clone,
impl<Hist, R, E, S, Res, Energy> GlueAble<Hist> for WangLandau1T<Hist, R, E, S, Res, Energy>where
Hist: Clone,
fn push_glue_entry_ignoring( &self, job: &mut GlueJob<Hist>, ignore_idx: &[usize] )
fn push_glue_entry(&self, job: &mut GlueJob<H>)
source§impl<Hist, Rng, Ensemble, S, Res, Energy> Serialize for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
impl<Hist, Rng, Ensemble, S, Res, Energy> Serialize for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
source§impl<Hist, R, E, S, Res, Energy> TryFrom<WangLandau1T<Hist, R, E, S, Res, Energy>> for EntropicSampling<Hist, R, E, S, Res, Energy>
impl<Hist, R, E, S, Res, Energy> TryFrom<WangLandau1T<Hist, R, E, S, Res, Energy>> for EntropicSampling<Hist, R, E, S, Res, Energy>
§type Error = EntropicErrors
type Error = EntropicErrors
The type returned in the event of a conversion error.
source§impl<Hist, R, E, S, Res, Energy> WangLandau for WangLandau1T<Hist, R, E, S, Res, Energy>
impl<Hist, R, E, S, Res, Energy> WangLandau for WangLandau1T<Hist, R, E, S, Res, Energy>
source§fn log_f_threshold(&self) -> f64
fn log_f_threshold(&self) -> f64
returns currently set threshold for log_f Read more
source§fn set_log_f_threshold(
&mut self,
log_f_threshold: f64
) -> Result<f64, WangLandauErrors>
fn set_log_f_threshold( &mut self, log_f_threshold: f64 ) -> Result<f64, WangLandauErrors>
Try to set the threshold. Read more
source§fn write_log<W: Write>(&self, writer: W) -> Result<(), Error>
fn write_log<W: Write>(&self, writer: W) -> Result<(), Error>
Writes Information about the simulation to a file.
E.g. How many steps were performed.
source§fn mode(&self) -> WangLandauMode
fn mode(&self) -> WangLandauMode
Returns current wang landau mode Read more
source§fn step_counter(&self) -> usize
fn step_counter(&self) -> usize
Counter Read more
source§fn total_steps_rejected(&self) -> usize
fn total_steps_rejected(&self) -> usize
How many steps were rejected until now? Read more
source§fn total_steps_accepted(&self) -> usize
fn total_steps_accepted(&self) -> usize
How many steps were accepted until now? Read more
source§fn is_finished(&self) -> bool
fn is_finished(&self) -> bool
Checks wang landau threshold Read more
source§fn log_density_base10(&self) -> Vec<f64>
fn log_density_base10(&self) -> Vec<f64>
Current (non normalized) estimate of log10(P(E)) Read more
source§fn log_density_base(&self, base: f64) -> Vec<f64>
fn log_density_base(&self, base: f64) -> Vec<f64>
Current (non normalized) estimate of log_base(P(E)) Read more
source§fn steps_total(&self) -> usize
fn steps_total(&self) -> usize
Counter Read more
source§fn fraction_accepted_total(&self) -> f64
fn fraction_accepted_total(&self) -> f64
Calculate, which fraction of steps were accepted Read more
source§fn fraction_rejected_total(&self) -> f64
fn fraction_rejected_total(&self) -> f64
Calculate, which fraction of steps were rejected Read more
source§impl<Hist, R, E, S, Res, Energy> WangLandauEnergy<Energy> for WangLandau1T<Hist, R, E, S, Res, Energy>
impl<Hist, R, E, S, Res, Energy> WangLandauEnergy<Energy> for WangLandau1T<Hist, R, E, S, Res, Energy>
source§impl<Hist, R, E, S, Res, Energy> WangLandauEnsemble<E> for WangLandau1T<Hist, R, E, S, Res, Energy>
impl<Hist, R, E, S, Res, Energy> WangLandauEnsemble<E> for WangLandau1T<Hist, R, E, S, Res, Energy>
source§impl<Hist, R, E, S, Res, Energy> WangLandauHist<Hist> for WangLandau1T<Hist, R, E, S, Res, Energy>
impl<Hist, R, E, S, Res, Energy> WangLandauHist<Hist> for WangLandau1T<Hist, R, E, S, Res, Energy>
Auto Trait Implementations§
impl<Hist, Rng, Ensemble, S, Res, Energy> RefUnwindSafe for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>where
Energy: RefUnwindSafe,
Ensemble: RefUnwindSafe,
Hist: RefUnwindSafe,
Res: RefUnwindSafe,
Rng: RefUnwindSafe,
S: RefUnwindSafe,
impl<Hist, Rng, Ensemble, S, Res, Energy> Send for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
impl<Hist, Rng, Ensemble, S, Res, Energy> Sync for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
impl<Hist, Rng, Ensemble, S, Res, Energy> Unpin for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>
impl<Hist, Rng, Ensemble, S, Res, Energy> UnwindSafe for WangLandau1T<Hist, Rng, Ensemble, S, Res, Energy>where
Energy: UnwindSafe,
Ensemble: UnwindSafe,
Hist: UnwindSafe,
Res: UnwindSafe,
Rng: UnwindSafe,
S: UnwindSafe,
Blanket Implementations§
source§impl<T> BorrowMut<T> for Twhere
T: ?Sized,
impl<T> BorrowMut<T> for Twhere
T: ?Sized,
source§fn borrow_mut(&mut self) -> &mut T
fn borrow_mut(&mut self) -> &mut T
Mutably borrows from an owned value. Read more
§impl<S, T> Cast<T> for Swhere
T: Conv<S>,
impl<S, T> Cast<T> for Swhere
T: Conv<S>,
§impl<S, T> CastApprox<T> for Swhere
T: ConvApprox<S>,
impl<S, T> CastApprox<T> for Swhere
T: ConvApprox<S>,
§fn try_cast_approx(self) -> Result<T, Error>
fn try_cast_approx(self) -> Result<T, Error>
§fn cast_approx(self) -> T
fn cast_approx(self) -> T
§impl<S, T> CastFloat<T> for Swhere
T: ConvFloat<S>,
impl<S, T> CastFloat<T> for Swhere
T: ConvFloat<S>,
§fn cast_trunc(self) -> T
fn cast_trunc(self) -> T
Cast to integer, truncating Read more
§fn cast_nearest(self) -> T
fn cast_nearest(self) -> T
Cast to the nearest integer Read more
§fn cast_floor(self) -> T
fn cast_floor(self) -> T
Cast the floor to an integer Read more
§fn try_cast_trunc(self) -> Result<T, Error>
fn try_cast_trunc(self) -> Result<T, Error>
Try converting to integer with truncation Read more
§fn try_cast_nearest(self) -> Result<T, Error>
fn try_cast_nearest(self) -> Result<T, Error>
Try converting to the nearest integer Read more
§fn try_cast_floor(self) -> Result<T, Error>
fn try_cast_floor(self) -> Result<T, Error>
Try converting the floor to an integer Read more
§fn try_cast_ceil(self) -> Result<T, Error>
fn try_cast_ceil(self) -> Result<T, Error>
Try convert the ceiling to an integer Read more