Commit 15e3f5cd authored by Alberto Ramos's avatar Alberto Ramos

Adds supports for gaps in the MC histories

parent b2fe6cd2
......@@ -27,6 +27,7 @@ robust and **exact error propagation even in iterative algorithms**.
different parameters).
- Support for **replicas** (i.e. several runs with the same simulation
parameters).
- Irrelagular MC measurements are handled transparently.
- Standalone **portable** implementation without any external
dependencies.
- **Fast** computation of autocorrelation functions with the [FFT
......
......@@ -813,6 +813,44 @@ they cannot be called on derived observables.
\subsection{Externally accessible functions}
\subsubsection{\texttt{aderr\_pobs(data, id\_msm, ivrep, nmsm)}}
Returns a primary onservable with the measurements contained in
\texttt{data(:)} corresponding to MC time \texttt{id\_msm(:)}. The
total number of measurements in the ensemble and the replica vector
are specified in \texttt{nmsm} and \texttt{id\_msm(:)} respectively.
Note that this call admits gaps in the MC history. For example if some
ensemble has measurements at MC times $1,2,\dots,100$ for most
observables, but some observable has only been measured at MC times
$1,3,5,\dots,99$, this data can still be used in any expression by
just calling
\begin{verbatim}
x = aderr_pobs(data(1:50),[(2*i-1,i=1,50)], [100], 100)
\end{verbatim}
will load in \texttt{x} the measurements stored in \texttt{data} that
are \emph{only available for odd values of the MC time}.
Completely irregular MC histories are transparently dealt with in the code.
\begin{description}
\item[\texttt{data(:)}:] Type \texttt{real double precision} one
dimensional array. The MC data.
\item[\texttt{id\_msm(:)}:] Type \texttt{integer} one dimensional
array. The measurements ID (i.e. ``trajectories'' in lattice jargon)
of each of the measurements.
\item[\texttt{ivrep(:)}:] Type \texttt{integer} one simansional
array. The replica vector of the ensemble.
\item[\texttt{nmsm}:] Type \texttt{integer}. The total number of
measurements in the ensemble.
\item[Output: ] Type \texttt{uwreal}. The primary observable.
\end{description}
\subsubsection{\texttt{aderr\_new\_id(name, [id])}}
Stores \texttt{name} in the database and associate it with ID
......
......@@ -254,6 +254,12 @@ module aderrors
type (uwreal), intent (in) :: a(:)
real (kind=DP), intent (in) :: der(size(x),size(a)), mns(size(x))
end subroutine
module subroutine addobs_ds(x, a, der, mns)
type (uwreal), intent (inout) :: x
type (uwreal), intent (in) :: a(:)
real (kind=DP), intent (in) :: der(:), mns
end subroutine
end interface
interface assignment (=)
......@@ -296,14 +302,15 @@ module aderrors
module procedure uwdobs_f, uwdobs_s
end interface uwdobs
interface addobs
module procedure addobs_f, addobs_s, addobs_d
module procedure addobs_f, addobs_s, addobs_d, addobs_ds
end interface addobs
public :: uwdobs, addobs
interface
module subroutine uwerr(p)
module subroutine uwerr(p, ids_texp)
class (uwreal), intent (inout) :: p
integer, intent (in), optional :: ids_texp(:)
end subroutine
end interface
......@@ -604,7 +611,8 @@ module aderrors
end function
end interface
public :: aderr_set_mode, aderr_get_mode, adset_default_dsig, adset_default_stau
public :: aderr_set_mode, aderr_get_mode, adset_default_dsig, &
adset_default_stau, aderr_pobs
contains
......@@ -1247,7 +1255,33 @@ contains
return
end subroutine print_hist
! *****************************
! *
function aderr_pobs(data, id_msm, ivrep, nmsm) result (p)
! *
! *****************************
type (uwreal) :: p
real (kind=DP), intent (in) :: data(:)
integer, intent (in) :: id_msm(:), ivrep(:), nmsm
integer :: i
if (size(data).ne.size(id_msm)) call module_error('[aderr_pobs]',&
' Data and trajectories do not match')
call p%init(1, [nmsm], [size(ivrep)], ivrep)
p%mean = sum(data)/real(size(data),kind=DP)
p%data = 0.0_DP
do i = 1, size(id_msm)
p%data(id_msm(i)) = data(i) - p%mean
end do
p%data = real(nmsm,kind=DP)/real(size(id_msm)) * p%data
return
end function aderr_pobs
! ********************************
! *
Subroutine module_error(routine, msg)
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment