Merge pull request #4434 from akohlmey/collected-small-changes

Collected small changes and fixes

.github/workflows/coverity.yml

@@ -67,7 +67,6 @@ jobs:
         -D PKG_MANIFOLD=on \
         -D PKG_MDI=on \
         -D PKG_MGPT=on \
-        -D PKG_ML-PACE=on \
         -D PKG_ML-RANN=on \
         -D PKG_MOLFILE=on \
         -D PKG_NETCDF=on \

.github/workflows/lammps-gui-flatpak.yml

@@ -0,0 +1,53 @@
+# GitHub action to build LAMMPS-GUI as a flatpak bundle
+name: "Build LAMMPS-GUI as flatpak bundle"
+
+on:
+  push:
+    branches:
+      - develop
+
+  workflow_dispatch:
+
+concurrency:
+  group: ${{ github.event_name }}-${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: ${{github.event_name == 'pull_request'}}
+
+jobs:
+  build:
+    name: LAMMPS-GUI flatpak build
+    if: ${{ github.repository == 'lammps/lammps' }}
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v4
+      with:
+        fetch-depth: 2
+
+    - name: Install extra packages
+      run: |
+        sudo apt-get update
+        sudo apt-get install -y ccache \
+                                libeigen3-dev \
+                                libcurl4-openssl-dev \
+                                mold \
+                                ninja-build \
+                                python3-dev \
+                                flatpak \
+                                flatpak-builder
+
+    - name: Set up access to flatpak repo
+      run: flatpak --user remote-add --if-not-exists flathub https://dl.flathub.org/repo/flathub.flatpakrepo
+
+    - name: Build flatpak
+      run: |
+        mkdir flatpack-state
+        sed -i -e 's/branch:.*/branch: develop/' tools/lammps-gui/org.lammps.lammps-gui.yml
+        flatpak-builder --force-clean --verbose --repo=flatpak-repo \
+                               --install-deps-from=flathub --state-dir=flatpak-state \
+                               --user --ccache --default-branch=${{ github.ref_name }} \
+                               flatpak-build tools/lammps-gui/org.lammps.lammps-gui.yml
+        flatpak build-bundle --runtime-repo=https://flathub.org/repo/flathub.flatpakrepo \
+                               --verbose  flatpak-repo LAMMPS-Linux-x86_64-GUI.flatpak \
+                               org.lammps.lammps-gui ${{ github.ref_name }}
+        flatpak install -y -v --user LAMMPS-Linux-x86_64-GUI.flatpak

doc/src/Commands_removed.rst

@@ -1,6 +1,10 @@
 Removed commands and packages
 =============================
 
+.. contents::
+
+------
+
 This page lists LAMMPS commands and packages that have been removed from
 the distribution and provides suggestions for alternatives or
 replacements.  LAMMPS has special dummy styles implemented, that will
@@ -8,47 +12,60 @@ stop LAMMPS and print a suitable error message in most cases, when a
 style/command is used that has been removed or will replace the command
 with the direct alternative (if available) and print a warning.
 
-restart2data tool
+LAMMPS shell
------------------
+------------
 
-.. versionchanged:: 23Nov2013
+.. versionchanged:: 29Aug2024
 
-The functionality of the restart2data tool has been folded into the
+The LAMMPS shell has been removed from the LAMMPS distribution. Users
-LAMMPS executable directly instead of having a separate tool.  A
+are encouraged to use the :ref:`LAMMPS-GUI <lammps_gui>` tool instead.
-combination of the commands :doc:`read_restart <read_restart>` and
-:doc:`write_data <write_data>` can be used to the same effect.  For
-added convenience this conversion can also be triggered by
-:doc:`command-line flags <Run_options>`
 
-Fix ave/spatial and fix ave/spatial/sphere
+i-PI tool
-------------------------------------------
+---------
 
-.. deprecated:: 11Dec2015
+.. versionchanged:: 27Jun2024
 
-The fixes ave/spatial and ave/spatial/sphere have been removed from LAMMPS
+The i-PI tool has been removed from the LAMMPS distribution.  Instead,
-since they were superseded by the more general and extensible "chunk
+instructions to install i-PI from PyPI via pip are provided.
-infrastructure".  Here the system is partitioned in one of many possible
-ways through the :doc:`compute chunk/atom <compute_chunk_atom>` command
-and then averaging is done using :doc:`fix ave/chunk <fix_ave_chunk>`.
-Please refer to the :doc:`chunk HOWTO <Howto_chunk>` section for an overview.
 
-Box command
+USER-REAXC package
------------
+------------------
 
-.. deprecated:: 22Dec2022
+.. deprecated:: 7Feb2024
 
-The *box* command has been removed and the LAMMPS code changed so it won't
+The USER-REAXC package has been renamed to :ref:`REAXFF <PKG-REAXFF>`.
-be needed.  If present, LAMMPS will ignore the command and print a warning.
+In the process also the pair style and related fixes were renamed to use
+the "reaxff" string instead of "reax/c". For a while LAMMPS was maintaining
+backward compatibility by providing aliases for the styles.  These have
+been removed, so using "reaxff" is now *required*.
 
-Reset_ids, reset_atom_ids, reset_mol_ids commands
+MPIIO package
--------------------------------------------------
+-------------
 
-.. deprecated:: 22Dec2022
+.. deprecated:: 21Nov2023
 
-The *reset_ids*, *reset_atom_ids*, and *reset_mol_ids* commands have
+The MPIIO package has been removed from LAMMPS since it was unmaintained
-been folded into the :doc:`reset_atoms <reset_atoms>` command.  If
+for many years and thus not updated to incorporate required changes that
-present, LAMMPS will replace the commands accordingly and print a
+had been applied to the corresponding non-MPIIO commands. As a
-warning.
+consequence the MPIIO commands had become unreliable and sometimes
+crashing LAMMPS or corrupting data.  Similar functionality is available
+through the :ref:`ADIOS package <PKG-ADIOS>` and the :ref:`NETCDF
+package <PKG-NETCDF>`.  Also, the :doc:`dump_modify nfile or dump_modify
+fileper <dump_modify>` keywords may be used for an efficient way of
+writing out dump files when running on large numbers of processors.
+Similarly, the "nfile" and "fileper" keywords exist for restarts:
+see :doc:`restart <restart>`, :doc:`read_restart <read_restart>`,
+:doc:`write_restart <write_restart>`.
+
+MSCG package
+------------
+
+.. deprecated:: 21Nov2023
+
+The MSCG package has been removed from LAMMPS since it was unmaintained
+for many years and instead superseded by the `OpenMSCG software
+<https://software.rcc.uchicago.edu/mscg/>`_ of the Voth group at the
+University of Chicago, which can be used independent from LAMMPS.
 
 LATTE package
 -------------
@@ -64,18 +81,6 @@ packages, including LATTE.  See the ``examples/QUANTUM`` dir and the
 with LATTE as a plugin library (similar to the way fix latte worked), as
 well as on a different set of MPI processors.
 
-MEAM package
-------------
-
-The MEAM package in Fortran has been replaced by a C++ implementation.
-The code in the :ref:`MEAM package <PKG-MEAM>` is a translation of the
-Fortran code of MEAM into C++, which removes several restrictions
-(e.g. there can be multiple instances in hybrid pair styles) and allows
-for some optimizations leading to better performance.  The pair style
-:doc:`meam <pair_meam>` has the exact same syntax.  For a transition
-period the C++ version of MEAM was called USER-MEAMC so it could
-coexist with the Fortran version.
-
 Minimize style fire/old
 -----------------------
 
@@ -97,38 +102,38 @@ The same functionality is available through
 :doc:`bond style mesocnt <bond_mesocnt>` and
 :doc:`angle style mesocnt <angle_mesocnt>`.
 
-MPIIO package
+Box command
--------------
+-----------
 
-.. deprecated:: 21Nov2023
+.. deprecated:: 22Dec2022
 
-The MPIIO package has been removed from LAMMPS since it was unmaintained
+The *box* command has been removed and the LAMMPS code changed so it won't
-for many years and thus not updated to incorporate required changes that
+be needed.  If present, LAMMPS will ignore the command and print a warning.
-had been applied to the corresponding non-MPIIO commands. As a
-consequence the MPIIO commands had become unreliable and sometimes
-crashing LAMMPS or corrupting data.  Similar functionality is available
-through the :ref:`ADIOS package <PKG-ADIOS>` and the :ref:`NETCDF
-package <PKG-NETCDF>`.  Also, the :doc:`dump_modify nfile or dump_modify
-fileper <dump_modify>` keywords may be used for an efficient way of
-writing out dump files when running on large numbers of processors.
-Similarly, the "nfile" and "fileper" keywords exist for restarts:
-see :doc:`restart <restart>`, :doc:`read_restart <read_restart>`,
-:doc:`write_restart <write_restart>`.
 
+Reset_ids, reset_atom_ids, reset_mol_ids commands
+-------------------------------------------------
 
-MSCG package
+.. deprecated:: 22Dec2022
-------------
 
-.. deprecated:: 21Nov2023
+The *reset_ids*, *reset_atom_ids*, and *reset_mol_ids* commands have
+been folded into the :doc:`reset_atoms <reset_atoms>` command.  If
+present, LAMMPS will replace the commands accordingly and print a
+warning.
 
-The MSCG package has been removed from LAMMPS since it was unmaintained
+MESSAGE package
-for many years and instead superseded by the `OpenMSCG software
+---------------
-<https://software.rcc.uchicago.edu/mscg/>`_ of the Voth group at the
+
-University of Chicago, which can be used independent from LAMMPS.
+.. deprecated:: 4May2022
+
+The MESSAGE package has been removed since it was superseded by the
+:ref:`MDI package <PKG-MDI>`. MDI implements the same functionality
+and in a more general way with direct support for more applications.
 
 REAX package
 ------------
 
+.. deprecated:: 4Jan2019
+
 The REAX package has been removed since it was superseded by the
 :ref:`REAXFF package <PKG-REAXFF>`.  The REAXFF package has been tested
 to yield equivalent results to the REAX package, offers better
@@ -138,20 +143,25 @@ syntax compatible with the removed reax pair style, so input files will
 have to be adapted.  The REAXFF package was originally called
 USER-REAXC.
 
-USER-REAXC package
+MEAM package
-------------------
+------------
 
-.. deprecated:: 7Feb2024
+.. deprecated:: 4Jan2019
 
-The USER-REAXC package has been renamed to :ref:`REAXFF <PKG-REAXFF>`.
+The MEAM package in Fortran has been replaced by a C++ implementation.
-In the process also the pair style and related fixes were renamed to use
+The code in the :ref:`MEAM package <PKG-MEAM>` is a translation of the
-the "reaxff" string instead of "reax/c". For a while LAMMPS was maintaining
+Fortran code of MEAM into C++, which removes several restrictions
-backward compatibility by providing aliases for the styles.  These have
+(e.g. there can be multiple instances in hybrid pair styles) and allows
-been removed, so using "reaxff" is now *required*.
+for some optimizations leading to better performance.  The pair style
+:doc:`meam <pair_meam>` has the exact same syntax.  For a transition
+period the C++ version of MEAM was called USER-MEAMC so it could
+coexist with the Fortran version.
 
 USER-CUDA package
 -----------------
 
+.. deprecated:: 31May2016
+
 The USER-CUDA package had been removed, since it had been unmaintained
 for a long time and had known bugs and problems.  Significant parts of
 the design were transferred to the
@@ -160,19 +170,27 @@ performance characteristics on NVIDIA GPUs. Both, the KOKKOS
 and the :ref:`GPU package <PKG-GPU>` are maintained
 and allow running LAMMPS with GPU acceleration.
 
-i-PI tool
+Fix ave/spatial and fix ave/spatial/sphere
----------
+------------------------------------------
 
-.. versionchanged:: 27Jun2024
+.. deprecated:: 11Dec2015
 
-The i-PI tool has been removed from the LAMMPS distribution.  Instead,
+The fixes ave/spatial and ave/spatial/sphere have been removed from LAMMPS
-instructions to install i-PI from PyPI via pip are provided.
+since they were superseded by the more general and extensible "chunk
+infrastructure".  Here the system is partitioned in one of many possible
+ways through the :doc:`compute chunk/atom <compute_chunk_atom>` command
+and then averaging is done using :doc:`fix ave/chunk <fix_ave_chunk>`.
+Please refer to the :doc:`chunk HOWTO <Howto_chunk>` section for an overview.
 
-LAMMPS shell
+restart2data tool
-------------
+-----------------
 
-.. versionchanged:: 29Aug2024
+.. deprecated:: 23Nov2013
 
-The LAMMPS shell has been removed from the LAMMPS distribution. Users
+The functionality of the restart2data tool has been folded into the
-are encouraged to use the :ref:`LAMMPS-GUI <lammps_gui>` tool instead.
+LAMMPS executable directly instead of having a separate tool.  A
+combination of the commands :doc:`read_restart <read_restart>` and
+:doc:`write_data <write_data>` can be used to the same effect.  For
+added convenience this conversion can also be triggered by
+:doc:`command-line flags <Run_options>`
 

doc/src/Fortran.rst

@@ -321,6 +321,8 @@ of the contents of the :f:mod:`LIBLAMMPS` Fortran interface to LAMMPS.
    :ftype set_string_variable: subroutine
    :f set_internal_variable: :f:subr:`set_internal_variable`
    :ftype set_internal_variable: subroutine
+   :f eval: :f:func:`eval`
+   :ftype eval: function
    :f gather_atoms: :f:subr:`gather_atoms`
    :ftype gather_atoms: subroutine
    :f gather_atoms_concat: :f:subr:`gather_atoms_concat`

doc/src/Howto_lammps_gui.rst

@@ -64,13 +64,18 @@ simple LAMMPS simulations.  It is very suitable for tutorials on LAMMPS
 since you only need to learn how to use a single program for most tasks
 and thus time can be saved and people can focus on learning LAMMPS.
 The tutorials at https://lammpstutorials.github.io/ are specifically
-updated for use with LAMMPS-GUI.
+updated for use with LAMMPS-GUI and can their tutorial materials can
+be downloaded and loaded directly from the GUI.
 
 Another design goal is to keep the barrier low when replacing part of
 the functionality of LAMMPS-GUI with external tools.  That said, LAMMPS-GUI
 has some unique functionality that is not found elsewhere:
 
 - auto-adapting to features available in the integrated LAMMPS library
+- auto-completion for LAMMPS commands and options
+- context-sensitive online help
+- start and stop of simulations via mouse or keyboard
+- monitoring of simulation progress
 - interactive visualization using the :doc:`dump image <dump_image>`
   command with the option to copy-paste the resulting settings
 - automatic slide show generation from dump image out at runtime

doc/src/Library_execute.rst

@@ -7,6 +7,7 @@ This section documents the following functions:
 - :cpp:func:`lammps_command`
 - :cpp:func:`lammps_commands_list`
 - :cpp:func:`lammps_commands_string`
+- :cpp:func:`lammps_expand`
 
 --------------------
 
@@ -79,3 +80,8 @@ Below is a short example using some of these functions.
 .. doxygenfunction:: lammps_commands_string
    :project: progguide
 
+-----------------------
+
+.. doxygenfunction:: lammps_expand
+   :project: progguide
+

doc/src/Library_objects.rst

@@ -1,5 +1,5 @@
-Compute, fixes, variables
+Computes, fixes, variables
-=========================
+==========================
 
 This section documents accessing or modifying data stored by computes,
 fixes, or variables in LAMMPS using the following functions:
@@ -12,6 +12,7 @@ fixes, or variables in LAMMPS using the following functions:
 - :cpp:func:`lammps_set_string_variable`
 - :cpp:func:`lammps_set_internal_variable`
 - :cpp:func:`lammps_variable_info`
+- :cpp:func:`lammps_eval`
 
 -----------------------
 
@@ -55,6 +56,11 @@ fixes, or variables in LAMMPS using the following functions:
 
 -----------------------
 
+.. doxygenfunction:: lammps_eval
+   :project: progguide
+
+-----------------------
+
 .. doxygenenum:: _LMP_DATATYPE_CONST
 
 .. doxygenenum:: _LMP_STYLE_CONST

examples/COUPLE/plugin/liblammpsplugin.c

@@ -117,6 +117,7 @@ liblammpsplugin_t *liblammpsplugin_load(const char *lib)
   ADDSYM(set_string_variable);
   ADDSYM(set_internal_variable);
   ADDSYM(variable_info);
+  ADDSYM(eval);
 
   ADDSYM(gather_atoms);
   ADDSYM(gather_atoms_concat);

examples/COUPLE/plugin/liblammpsplugin.h

@@ -163,6 +163,7 @@ struct _liblammpsplugin {
   int (*set_string_variable)(void *, const char *, const char *);
   int (*set_internal_variable)(void *, const char *, double);
   int (*variable_info)(void *, int, char *, int);
+  double (*eval)(void *, const char *);
 
   void (*gather_atoms)(void *, const char *, int, int, void *);
   void (*gather_atoms_concat)(void *, const char *, int, int, void *);

fortran/lammps.f90

@@ -126,6 +126,7 @@ MODULE LIBLAMMPS
     PROCEDURE :: set_variable           => lmp_set_variable
     PROCEDURE :: set_string_variable    => lmp_set_string_variable
     PROCEDURE :: set_internal_variable  => lmp_set_internal_variable
+    PROCEDURE :: eval                   => lmp_eval
     PROCEDURE, PRIVATE :: lmp_gather_atoms_int
     PROCEDURE, PRIVATE :: lmp_gather_atoms_double
     GENERIC   :: gather_atoms           => lmp_gather_atoms_int, &
@@ -618,7 +619,14 @@ MODULE LIBLAMMPS
       INTEGER(c_int) :: lammps_set_internal_variable
     END FUNCTION lammps_set_internal_variable
 
-    SUBROUTINE lammps_gather_atoms(handle, name, type, count, data) BIND(C)
+    FUNCTION lammps_eval(handle, expr) BIND(C)
+      IMPORT :: c_ptr, c_double
+      IMPLICIT NONE
+      TYPE(c_ptr), VALUE :: handle, expr
+      REAL(c_double) :: lammps_eval
+    END FUNCTION lammps_eval
+
+    SUBROUTINE lammps_gather_atoms(handle, name, TYPE, count, DATA) BIND(C)
       IMPORT :: c_int, c_ptr
       IMPLICIT NONE
       TYPE(c_ptr), VALUE :: handle, name, data
@@ -1812,7 +1820,7 @@ CONTAINS
   SUBROUTINE lmp_set_internal_variable(self, name, val)
     CLASS(lammps), INTENT(IN) :: self
     CHARACTER(LEN=*), INTENT(IN) :: name
-    REAL(KIND=c_double), INTENT(IN) :: val
+    REAL(c_double), INTENT(IN) :: val
     INTEGER :: err
     TYPE(c_ptr) :: Cname
 
@@ -1826,6 +1834,18 @@ CONTAINS
     END IF
   END SUBROUTINE lmp_set_internal_variable
 
+  ! equivalent function to lammps_eval
+  FUNCTION lmp_eval(self, expr)
+    CLASS(lammps), INTENT(IN) :: self
+    CHARACTER(LEN=*), INTENT(IN) :: expr
+    REAL(c_double) :: lmp_eval
+    TYPE(c_ptr) :: Cexpr
+
+    Cexpr = f2c_string(expr)
+    lmp_eval = lammps_eval(self%handle, Cexpr)
+    CALL lammps_free(Cexpr)
+  END FUNCTION lmp_eval
+
   ! equivalent function to lammps_gather_atoms (for integers)
   SUBROUTINE lmp_gather_atoms_int(self, name, count, data)
     CLASS(lammps), INTENT(IN) :: self

python/lammps/core.py

@@ -103,7 +103,7 @@ class command_wrapper(object):
     This method is where the Python 'magic' happens. If a method is not
     defined by the class command_wrapper, it assumes it is a LAMMPS command. It takes
     all the arguments, concatinates them to a single string, and executes it using
-    :py:meth:`lammps.command()`.
+    :py:meth:`lammps.command`.
 
     Starting with Python 3.6 it also supports keyword arguments. key=value is
     transformed into 'key value'. Note, since these have come last in the
@@ -422,6 +422,9 @@ class lammps(object):
     self.lib.lammps_extract_variable_datatype.argtypes = [c_void_p, c_char_p]
     self.lib.lammps_extract_variable_datatype.restype = c_int
 
+    self.lib.lammps_eval.argtypes = [c_void_p, c_char_p]
+    self.lib.lammps_eval.restype = c_double
+
     self.lib.lammps_fix_external_get_force.argtypes = [c_void_p, c_char_p]
     self.lib.lammps_fix_external_get_force.restype = POINTER(POINTER(c_double))
 
@@ -1671,6 +1674,30 @@ class lammps(object):
 
   # -------------------------------------------------------------------------
 
+  def eval(self, expr):
+    """ Evaluate a LAMMPS immediate variable expression
+
+    .. versionadded:: TBD
+
+    This function is a wrapper around the function :cpp:func:`lammps_eval`
+    of the C library interface.  It evaluates and expression like in
+    immediate variables and returns the value.
+
+    :param expr: immediate variable expression
+    :type name: string
+    :return: the result of the evaluation
+    :rtype: c_double
+    """
+
+    if expr: newexpr = expr.encode()
+    else: return None
+
+    with ExceptionCheck(self):
+      return self.lib.lammps_eval(self.lmp, newexpr)
+    return None
+
+  # -------------------------------------------------------------------------
+
   # return vector of atom properties gathered across procs
   # 3 variants to match src/library.cpp
   # name = atom property recognized by LAMMPS in atom->extract()

src/KOKKOS/compute_reaxff_atom_kokkos.cpp

@@ -63,6 +63,8 @@ void ComputeReaxFFAtomKokkos<DeviceType>::init()
 template<class DeviceType>
 void ComputeReaxFFAtomKokkos<DeviceType>::compute_bonds()
 {
+  invoked_bonds = update->ntimestep;
+
   if (atom->nmax > nmax) {
     memory->destroy(array_atom);
     nmax = atom->nmax;

src/QEQ/fix_qeq.cpp

@@ -41,6 +41,9 @@ using namespace LAMMPS_NS;
 using namespace FixConst;
 
 static constexpr double QSUMSMALL = 0.00001;
+static constexpr int MIN_CAP = 50;
+static constexpr double SAFE_ZONE = 1.2;
+static constexpr bigint MIN_NBRS = 100;
 
 namespace {
   class qeq_parser_error : public std::exception {

src/QEQ/fix_qeq.h

@@ -16,12 +16,6 @@
 
 #include "fix.h"
 
-#define EV_TO_KCAL_PER_MOL 14.4
-#define DANGER_ZONE 0.90
-#define MIN_CAP 50
-#define SAFE_ZONE 1.2
-#define MIN_NBRS 100
-
 namespace LAMMPS_NS {
 
 class FixQEq : public Fix {
@@ -36,6 +30,7 @@ class FixQEq : public Fix {
   void min_pre_force(int) override;
 
   double compute_scalar() override;
+  static constexpr double DANGER_ZONE = 0.90;
 
   // derived child classes must provide these functions
 

src/QEQ/fix_qeq_shielded.cpp

@@ -33,6 +33,8 @@
 
 using namespace LAMMPS_NS;
 
+static constexpr double EV_TO_KCAL_PER_MOL = 14.4;
+
 /* ---------------------------------------------------------------------- */
 
 FixQEqShielded::FixQEqShielded(LAMMPS *lmp, int narg, char **arg) : FixQEq(lmp, narg, arg)

src/finish.cpp

@@ -121,9 +121,9 @@ void Finish::end(int flag)
     MPI_Allreduce(&cpu_loop,&tmp,1,MPI_DOUBLE,MPI_SUM,world);
     cpu_loop = tmp/nprocs;
     if (time_loop > 0.0) cpu_loop = cpu_loop/time_loop*100.0;
+    output->thermo->footer();
 
     if (me == 0) {
-      output->thermo->footer();
       int ntasks = nprocs * nthreads;
       utils::logmesg(lmp,"Loop time of {:.6g} on {} procs for {} steps with {} atoms\n\n",
                      time_loop,ntasks,update->nsteps,atom->natoms);

src/library.cpp

@@ -516,6 +516,9 @@ treated as part of the command and will **not** start a second command.
 The function returns the expanded string in a new string buffer that
 must be freed with :cpp:func:`lammps_free` after use to avoid a memory leak.
 
+*See also*
+    :cpp:func:`lammps_eval`
+
 \endverbatim
  *
  * \param  handle  pointer to a previously created LAMMPS instance
@@ -850,7 +853,11 @@ This differs from :cpp:func:`lammps_get_thermo` in that it does **not**
 trigger an evaluation.  Instead it provides direct access to a read-only
 location of the last thermo output data and the corresponding keyword
 strings.  How to handle the return value depends on the value of the *what*
-argument string.
+argument string.  When accessing the data from a concurrent thread while
+LAMMPS is running, the cache needs to be locked first and then unlocked
+after the data is obtained, so that the data is not corrupted while
+reading in case LAMMPS wants to update it at the same time.  Outside
+of a run, the lock/unlock calls have no effect.
 
 .. note::
 
@@ -899,6 +906,14 @@ argument string.
      - actual field data for column
      - pointer to int, int64_t or double
      - yes
+   * - lock
+     - acquires lock to thermo data cache
+     - NULL pointer
+     - no
+   * - unlock
+     - releases lock to thermo data cache
+     - NULL pointer
+     - no
 
 \endverbatim
  *
@@ -954,8 +969,14 @@ void *lammps_last_thermo(void *handle, const char *what, int index)
       } else if (field.type == multitype::LAMMPS_DOUBLE) {
         val = (void *) &field.data.d;
       }
-
+    } else if (strcmp(what, "lock") == 0) {
+      th->lock_cache();
+      val = nullptr;
+    } else if (strcmp(what, "unlock") == 0) {
+      th->unlock_cache();
+      val = nullptr;
     } else val = nullptr;
+
   }
   END_CAPTURE
   return val;
@@ -2904,6 +2925,39 @@ int lammps_variable_info(void *handle, int idx, char *buffer, int buf_size) {
   return 0;
 }
 
+/** Evaluate an immediate variable expression
+ *
+\verbatim embed:rst
+
+.. versionadded:: TBD
+
+This function takes a string with an expression that can be used
+for :doc:`equal style variables <variable>`, evaluates it and returns
+the resulting (scalar) value as a floating point number.
+
+*See also*
+    :cpp:func:`lammps_expand`
+
+\endverbatim
+
+ * \param handle   pointer to a previously created LAMMPS instance cast to ``void *``.
+ * \param expr     string with expression
+ * \return         result from expression */
+
+double lammps_eval(void *handle, const char *expr)
+{
+  auto lmp = (LAMMPS *) handle;
+  double result = 0.0;
+
+  BEGIN_CAPTURE
+  {
+    result = lmp->input->variable->compute_equal(expr);
+  }
+  END_CAPTURE
+
+  return result;
+}
+
 // ----------------------------------------------------------------------
 // Library functions for scatter/gather operations of data
 // ----------------------------------------------------------------------

src/library.h

@@ -189,6 +189,7 @@ int lammps_set_variable(void *handle, const char *name, const char *str);
 int lammps_set_string_variable(void *handle, const char *name, const char *str);
 int lammps_set_internal_variable(void *handle, const char *name, double value);
 int lammps_variable_info(void *handle, int idx, char *buf, int bufsize);
+double lammps_eval(void *handle, const char *expr);
 
 /* ----------------------------------------------------------------------
  * Library functions for scatter/gather operations of data

src/thermo.cpp

@@ -100,8 +100,8 @@ static char fmtbuf[512];
 /* ---------------------------------------------------------------------- */
 
 Thermo::Thermo(LAMMPS *_lmp, int narg, char **arg) :
-    Pointers(_lmp), style(nullptr), vtype(nullptr), field2index(nullptr), argindex1(nullptr),
+    Pointers(_lmp), style(nullptr), vtype(nullptr), cache_mutex(nullptr), field2index(nullptr),
-    argindex2(nullptr), temperature(nullptr), pressure(nullptr), pe(nullptr)
+    argindex1(nullptr), argindex2(nullptr), temperature(nullptr), pressure(nullptr), pe(nullptr)
 {
   style = utils::strdup(arg[0]);
 
@@ -208,6 +208,7 @@ void Thermo::init()
   ValueTokenizer *format_line = nullptr;
   if (format_line_user.size()) format_line = new ValueTokenizer(format_line_user);
 
+  lock_cache();
   field_data.clear();
   field_data.resize(nfield);
   std::string format_this, format_line_user_def;
@@ -277,6 +278,7 @@ void Thermo::init()
         format[i] += fmt::format("{:<8} = {} ", keyword[i], format_this);
     }
   }
+  unlock_cache();
 
   // chop off trailing blank or add closing bracket if needed and then add newline
   if (lineflag == ONELINE)
@@ -320,6 +322,9 @@ void Thermo::init()
   if (index_press_scalar >= 0) pressure = computes[index_press_scalar];
   if (index_press_vector >= 0) pressure = computes[index_press_vector];
   if (index_pe >= 0) pe = computes[index_pe];
+
+  // create mutex to protect access to cached thermo data
+  cache_mutex = new std::mutex;
 }
 
 /* ----------------------------------------------------------------------
@@ -366,11 +371,19 @@ void Thermo::header()
 
 /* ---------------------------------------------------------------------- */
 
+// called at the end of a run from Finish class
+
 void Thermo::footer()
 {
+  if (comm->me == 0) {
     if (lineflag == YAMLLINE) utils::logmesg(lmp, "...\n");
   }
 
+  // no more locking for cached thermo data access needed
+  delete cache_mutex;
+  cache_mutex = nullptr;
+}
+
 /* ---------------------------------------------------------------------- */
 
 void Thermo::compute(int flag)
@@ -422,6 +435,7 @@ void Thermo::compute(int flag)
   }
 
   // add each thermo value to line with its specific format
+  lock_cache();
   field_data.clear();
   field_data.resize(nfield);
 
@@ -441,6 +455,7 @@ void Thermo::compute(int flag)
       field_data[ifield] = bivalue;
     }
   }
+  unlock_cache();
 
   // print line to screen and logfile
 
@@ -579,7 +594,8 @@ void Thermo::modify_params(int narg, char **arg)
       if (iarg + 2 > narg) error->all(FLERR, "Illegal thermo_modify command");
       triclinic_general = utils::logical(FLERR, arg[iarg + 1], false, lmp);
       if (triclinic_general && !domain->triclinic_general)
-        error->all(FLERR,"Thermo_modify triclinic/general cannot be used "
+        error->all(FLERR,
+                   "Thermo_modify triclinic/general cannot be used "
                    "if simulation box is not general triclinic");
       iarg += 2;
 
@@ -1566,6 +1582,26 @@ int Thermo::evaluate_keyword(const std::string &word, double *answer)
   return 0;
 }
 
+/* ---------------------------------------------------------------------- */
+
+// lock cache for current thermo data
+
+void Thermo::lock_cache()
+{
+  // no locking outside of a run
+  if (!cache_mutex) return;
+  cache_mutex->lock();
+}
+
+// unlock cache for current thermo data
+
+void Thermo::unlock_cache()
+{
+  // no locking outside of a run
+  if (!cache_mutex) return;
+  cache_mutex->unlock();
+}
+
 /* ----------------------------------------------------------------------
    extraction of Compute, Fix, Variable results
    compute/fix are normalized by atoms if returning extensive value

src/thermo.h

@@ -16,6 +16,7 @@
 
 #include "pointers.h"
 #include <map>
+#include <mutex>
 
 namespace LAMMPS_NS {
 
@@ -43,6 +44,8 @@ class Thermo : protected Pointers {
   int evaluate_keyword(const std::string &, double *);
 
   // for accessing cached thermo and related data
+  void lock_cache();
+  void unlock_cache();
   const int *get_line() const { return &nline; }
   const char *get_image_fname() const { return image_fname.c_str(); }
 
@@ -82,6 +85,9 @@ class Thermo : protected Pointers {
   int nline;
   std::string image_fname;
 
+  // mutex for locking the cache
+  std::mutex *cache_mutex;
+
   // data used by routines that compute single values
 
   int ivalue;          // integer value to print

src/variable.cpp

@@ -457,15 +457,21 @@ void Variable::set(int narg, char **arg)
   // data = 2 values, 1st is string to eval, 2nd is filled on retrieval
 
   } else if (strcmp(arg[1],"equal") == 0) {
-    if (narg != 3)
+    if (narg < 3) utils::missing_cmd_args(FLERR, "variable equal", error);
-      error->all(FLERR,"Illegal variable command: expected 3 arguments but found {}{}",
+
-                 narg, utils::errorurl(3));
+    // combine excess arguments into single string with a blank as separator
+    std::string combined = arg[2];
+    for (int iarg = 3; iarg < narg; ++iarg) {
+      combined += ' ';
+      combined += arg[iarg];
+    }
+
     int ivar = find(arg[0]);
     if (ivar >= 0) {
       if (style[ivar] != EQUAL)
         error->all(FLERR,"Cannot redefine variable as a different style");
       delete[] data[ivar][0];
-      data[ivar][0] = utils::strdup(arg[2]);
+      data[ivar][0] = utils::strdup(combined);
       replaceflag = 1;
     } else {
       if (nvar == maxvar) grow();
@@ -474,7 +480,7 @@ void Variable::set(int narg, char **arg)
       which[nvar] = 0;
       pad[nvar] = 0;
       data[nvar] = new char*[num[nvar]];
-      data[nvar][0] = utils::strdup(arg[2]);
+      data[nvar][0] = utils::strdup(combined);
       data[nvar][1] = new char[VALUELENGTH];
       strcpy(data[nvar][1],"(undefined)");
     }
@@ -485,15 +491,21 @@ void Variable::set(int narg, char **arg)
   // data = 1 value, string to eval
 
   } else if (strcmp(arg[1],"atom") == 0) {
-    if (narg != 3)
+    if (narg < 3) utils::missing_cmd_args(FLERR, "variable atom", error);
-      error->all(FLERR,"Illegal variable command: expected 3 arguments but found {}{}",
+
-                 narg, utils::errorurl(3));
+    // combine excess arguments into single string with a blank as separator
+    std::string combined = arg[2];
+    for (int iarg = 3; iarg < narg; ++iarg) {
+      combined += ' ';
+      combined += arg[iarg];
+    }
+
     int ivar = find(arg[0]);
     if (ivar >= 0) {
       if (style[ivar] != ATOM)
         error->all(FLERR,"Cannot redefine variable as a different style");
       delete[] data[ivar][0];
-      data[ivar][0] = utils::strdup(arg[2]);
+      data[ivar][0] = utils::strdup(combined);
       replaceflag = 1;
     } else {
       if (nvar == maxvar) grow();
@@ -502,7 +514,7 @@ void Variable::set(int narg, char **arg)
       which[nvar] = 0;
       pad[nvar] = 0;
       data[nvar] = new char*[num[nvar]];
-      data[nvar][0] = utils::strdup(arg[2]);
+      data[nvar][0] = utils::strdup(combined);
     }
 
   // VECTOR
@@ -513,16 +525,22 @@ void Variable::set(int narg, char **arg)
   //   immediately store it as N-length vector and set dynamic flag to 0
 
   } else if (strcmp(arg[1],"vector") == 0) {
-    if (narg != 3)
+    if (narg < 3) utils::missing_cmd_args(FLERR, "variable atom", error);
-      error->all(FLERR,"Illegal variable command: expected 3 arguments but found {}{}",
+
-                 narg, utils::errorurl(3));
+    // combine excess arguments into single string with a blank as separator
+    std::string combined = arg[2];
+    for (int iarg = 3; iarg < narg; ++iarg) {
+      combined += ' ';
+      combined += arg[iarg];
+    }
+
     int ivar = find(arg[0]);
     if (ivar >= 0) {
       if (style[ivar] != VECTOR)
         error->all(FLERR,"Cannot redefine variable as a different style");
       delete[] data[ivar][0];
       delete[] data[ivar][1];
-      data[ivar][0] = utils::strdup(arg[2]);
+      data[ivar][0] = utils::strdup(combined);
       if (data[ivar][0][0] != '[')
         vecs[ivar].dynamic = 1;
       else {
@@ -539,7 +557,7 @@ void Variable::set(int narg, char **arg)
       which[nvar] = 0;
       pad[nvar] = 0;
       data[nvar] = new char*[num[nvar]];
-      data[nvar][0] = utils::strdup(arg[2]);
+      data[nvar][0] = utils::strdup(combined);
       if (data[nvar][0][0] != '[') {
         vecs[nvar].dynamic = 1;
         data[nvar][1] = nullptr;

tools/lammps-gui/lammps-gui.appdata.xml

@@ -61,6 +61,8 @@
         Highlight warnings and error messages in Output window
         Make Tutorial wizards more compact
         Include download and compilation of WHAM software from Alan Grossfield
+        Add dialog to run WHAM directly from LAMMPS-GUI
+        Use mutex to avoid corruption of thermo data
       </description>
     </release>
     <release version="1.6.11" timestamp="1725080055">

tools/lammps-gui/lammpsgui.cpp

@@ -1015,6 +1015,7 @@ void LammpsGui::logupdate()
         void *ptr = lammps.last_thermo("setup", 0);
         if (ptr && *(int *)ptr) return;
 
+        lammps.last_thermo("lock", 0);
         ptr = lammps.last_thermo("num", 0);
         if (ptr) {
             int ncols = *(int *)ptr;
@@ -1066,6 +1067,7 @@ void LammpsGui::logupdate()
                 chartwindow->add_data(step, data, i);
             }
         }
+        lammps.last_thermo("unlock", 0);
     }
 
     // update list of available image file names

tools/lammps-gui/update-wham-2.0.11.patch

@@ -55,8 +55,79 @@ index 0000000..b4f0fe6
 +  TYPE DOC
 +  PERMISSIONS OWNER_READ GROUP_READ WORLD_READ
 +)
+diff --git a/nr/locate.c b/nr/locate.c
+index 9f92dc0..f3bf294 100644
+--- a/nr/locate.c
++++ b/nr/locate.c
+@@ -11,7 +11,7 @@ void locate(double xx[], int n, double x, int *j)
+ 	ascnd=(xx[n] > xx[0]);  // I think this makes it zero based
+ 	while (ju-jl > 1) {
+ 		jm=(ju+jl) >> 1;
+-		if (x > xx[jm] == ascnd)
++		if ((x > xx[jm]) == ascnd)
+ 			jl=jm;
+ 		else
+ 			ju=jm;
+diff --git a/wham-2d/histogram.c b/wham-2d/histogram.c
+index 1bd1329..b5d1c01 100644
+--- a/wham-2d/histogram.c
++++ b/wham-2d/histogram.c
+@@ -78,14 +78,14 @@ return hp;
+ }
+ 
+ /* Get a value from a histogram structure
+- * Given i and j, the indices into the global histogram, return 
++ * Given i and j, the indices into the global histogram, return
+  * the correct histogram value.  Since we only store range of histogram
+- * indices containing the nonzero values, we have to check the index value 
++ * indices containing the nonzero values, we have to check the index value
+  * against the range in the structure.
+  */
+ double get_histval(struct histogram *hist, int i, int j)
+ {
+-if ( (i < hist->first_x) || (i > hist->last_x) 
++if ( (i < hist->first_x) || (i > hist->last_x)
+    ||(j < hist->first_y) || (j > hist->last_y))
+     {
+     return 0.0;
+@@ -239,7 +239,7 @@ return error;
+ 
+ 
+ // Calculate the free energy, setting the minimum value to 0
+-void calc_free(double **free, double **prob, double kT, 
++void calc_free(double **free, double **prob, double kT,
+                int use_mask, int **mask)
+ {
+ int i,j;
+@@ -257,7 +257,14 @@ for (i=0; i<NUM_BINSx; i++)
+             }
+         else
+             {
+-            free[i][j] = -kT * log(prob[i][j]);
++            if (prob[i][j] > 0.0)
++                {
++                free[i][j] = -kT * log(prob[i][j]);
++                }
++            else
++                {
++                free[i][j] = 0.0;
++                }
+             if (free[i][j] < min)
+                 {
+                 min = free[i][j];
+@@ -321,8 +328,6 @@ return 0.5*(springx*(dx*dx) + springy*(dy*dy));
+ 
+ void calc_coor(int i, int j, double *coor)
+ {
+-coor[0] = HIST_MINx + BIN_WIDTHx*((double)i+0.5); 
+-coor[1] = HIST_MINy + BIN_WIDTHy*((double)j+0.5); 
++coor[0] = HIST_MINx + BIN_WIDTHx*((double)i+0.5);
++coor[1] = HIST_MINy + BIN_WIDTHy*((double)j+0.5);
+ }
+-
+-
 diff --git a/wham-2d/wham-2d.c b/wham-2d/wham-2d.c
-index fb6e059..2c5594f 100644
+index fb6e059..a6b8483 100644
 --- a/wham-2d/wham-2d.c
 +++ b/wham-2d/wham-2d.c
 @@ -25,7 +25,7 @@
@@ -77,6 +148,15 @@ index fb6e059..2c5594f 100644
  
  int main(int argc, char *argv[])
  {
+@@ -57,7 +57,7 @@ double sum;
+ int iteration;
+ int max_iteration = 100000;
+ int numpad;
+-int **mask;
++int **mask = NULL;
+ int use_mask;
+ 
+ cpu1 = ((double) clock())/CLOCKS_PER_SEC;
 @@ -76,6 +76,61 @@ for (i=0; i<argc; i++)
      }
  printf("\n");
@@ -139,6 +219,143 @@ index fb6e059..2c5594f 100644
  PERIODICx = parse_periodic(argv[1], &PERIODx);
  if (PERIODICx)
      {
+@@ -360,8 +415,8 @@ while ( ! iConverged || first)
+             for (j=0; j< NUM_BINSy; j++)
+                 {
+                 calc_coor(i,j,coor);
+-                printf("%f\t%f\t%f\t%f\n", coor[0], coor[1], free_ene[i][j],
+-                                        prob[i][j]);
++                if (prob[i][j] != 0.0)
++                    printf("%f\t%f\t%f\t%f\n", coor[0], coor[1], free_ene[i][j], prob[i][j]);
+                 }
+             }
+ 
+@@ -444,8 +499,9 @@ if (!FREEFILE)
+         for (j=0; j< NUM_BINSy; j++)
+             {
+             calc_coor(i,j,coor);
+-            printf("%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                                       free_ene[i][j], final_prob[i][j]);
++            if (prob[i][j] != 0.0)
++                printf("%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                       free_ene[i][j], final_prob[i][j]);
+             }
+         }
+     exit(errno);
+@@ -461,25 +517,28 @@ else
+         for (j=-numpad; j<0; j++)
+             {
+             calc_coor(i,j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                    free_ene[NUM_BINSx+i][NUM_BINSy+j],
+-                    final_prob[NUM_BINSx+i][NUM_BINSy+j]);
++            if (prob[NUM_BINSx+i][NUM_BINSy+j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[NUM_BINSx+i][NUM_BINSy+j],
++                        final_prob[NUM_BINSx+i][NUM_BINSy+j]);
+             }
+         // center values in y
+         for (j=0; j<NUM_BINSy; j++)
+             {
+             calc_coor(i,j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                           free_ene[NUM_BINSx+i][j],
+-                           final_prob[NUM_BINSx+i][j]);
++            if (prob[NUM_BINSx+i][j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[NUM_BINSx+i][j],
++                        final_prob[NUM_BINSx+i][j]);
+             }
+         // trailing padding values in y
+         for (j=0; j<numpad; j++)
+             {
+             calc_coor(i,NUM_BINSy+j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                           free_ene[NUM_BINSx+i][j],
+-                           final_prob[NUM_BINSx+i][j]);
++            if (prob[NUM_BINSx+i][j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[NUM_BINSx+i][j],
++                        final_prob[NUM_BINSx+i][j]);
+             }
+         fprintf(FREEFILE, "\n");
+         }
+@@ -490,25 +549,28 @@ else
+         for (j=-numpad; j<0; j++)
+             {
+             calc_coor(i,j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                           free_ene[i][NUM_BINSy+j],
+-                           final_prob[i][NUM_BINSy+j]);
++            if (prob[i][NUM_BINSy+j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[i][NUM_BINSy+j],
++                        final_prob[i][NUM_BINSy+j]);
+             }
+         // center values in y
+         for (j=0; j<NUM_BINSy; j++)
+             {
+             calc_coor(i,j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                             free_ene[i][j],
+-                             final_prob[i][j]);
++            if (prob[i][j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[i][j],
++                        final_prob[i][j]);
+             }
+         // trailing padding values in y
+         for (j=0; j<numpad; j++)
+             {
+             calc_coor(i,NUM_BINSy+j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                             free_ene[i][j],
+-                             final_prob[i][j]);
++            if (prob[i][j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[i][j],
++                        final_prob[i][j]);
+             }
+         fprintf(FREEFILE, "\n");
+         }
+@@ -519,25 +581,28 @@ else
+         for (j=-numpad; j<0; j++)
+             {
+             calc_coor(NUM_BINSx+i,j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                             free_ene[i][NUM_BINSy+j],
+-                             final_prob[i][NUM_BINSy+j]);
++            if (prob[i][NUM_BINSy+j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[i][NUM_BINSy+j],
++                        final_prob[i][NUM_BINSy+j]);
+             }
+         // center values in y
+         for (j=0; j<NUM_BINSy; j++)
+             {
+             calc_coor(NUM_BINSx+i,j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                             free_ene[i][j],
+-                             final_prob[i][j]);
++            if (prob[i][j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[i][j],
++                        final_prob[i][j]);
+             }
+         // trailing padding values in y
+         for (j=0; j<numpad; j++)
+             {
+             calc_coor(NUM_BINSx+i,NUM_BINSy+j,coor);
+-            fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
+-                             free_ene[i][j],
+-                             final_prob[i][j]);
++            if (prob[i][j] != 0.0)
++                fprintf(FREEFILE,"%f\t%f\t%f\t%f\n", coor[0], coor[1],
++                        free_ene[i][j],
++                        final_prob[i][j]);
+             }
+         fprintf(FREEFILE, "\n");
+         }
 diff --git a/wham-2d/wham-2d.h b/wham-2d/wham-2d.h
 index b17e4bd..5fc17ff 100644
 --- a/wham-2d/wham-2d.h
@@ -163,10 +380,67 @@ index b17e4bd..5fc17ff 100644
  
  
  // Value inserted for the free energy of masked values
+diff --git a/wham/histogram.c b/wham/histogram.c
+index bc52d74..635b39f 100644
+--- a/wham/histogram.c
++++ b/wham/histogram.c
+@@ -16,7 +16,7 @@
+ 
+ 
+ //  Allocate memory for a histogram
+-struct histogram *hist_alloc(int first, int last, int num_points, 
++struct histogram *hist_alloc(int first, int last, int num_points,
+                              int num_mc_samples)
+ {
+ struct histogram *hp;
+@@ -45,9 +45,9 @@ return hp;
+ }
+ 
+ /* Get a value from a histogram structure
+- * Given index, the index into the global histogram, return 
++ * Given index, the index into the global histogram, return
+  * the correct histogram value.  Since we only store range of histogram
+- * indices containing the nonzero values, we have to check the index value 
++ * indices containing the nonzero values, we have to check the index value
+  * against the range in the structure.
+  */
+ double get_histval(struct histogram *hist, int index)
+@@ -213,8 +213,16 @@ double min = 1e50;
+ bin_min = 0;
+ for (i=0; i<NUM_BINS; i++)
+     {
+-    free[i] = -kT * log(prob[i]);
+-    if (free[i] < min) 
++    if (prob[i] > 0.0)
++        {
++        free[i] = -kT * log(prob[i]);
++        }
++    else
++        {
++        free[i] = 0.0;
++        }
++
++    if (free[i] < min)
+         {
+         min = free[i];
+         bin_min = i;
+@@ -252,5 +260,5 @@ return 0.5*dx*dx*spring;
+ 
+ double calc_coor(int i)
+ {
+-return HIST_MIN + BIN_WIDTH*((double)i+0.5); 
++return HIST_MIN + BIN_WIDTH*((double)i+0.5);
+ }
 diff --git a/wham/wham.c b/wham/wham.c
-index 487871b..1496eed 100644
+index 487871b..edb8125 100644
 --- a/wham/wham.c
 +++ b/wham/wham.c
+@@ -1,4 +1,4 @@
+-/* 
++/*
+  * WHAM -- Weighted Histogram Analysis Method
+  *
+  * Reference 1: Computer Physics Communications, 91(1995) 275-282, Benoit Roux
 @@ -21,7 +21,7 @@
  #include "wham.h"
  
@@ -184,6 +458,15 @@ index 487871b..1496eed 100644
  
  int main(int argc, char *argv[])
  {
+@@ -41,7 +42,7 @@ int first;
+ int bin_min;
+ int have_energy;
+ char *freefile;
+-FILE *METAFILE, *FREEFILE; 
++FILE *METAFILE, *FREEFILE;
+ struct hist_group *hist_group;
+ struct histogram  *hp;
+ double coor;
 @@ -82,6 +83,61 @@ for (i=0; i<argc; i++)
      }
  printf("\n");
@@ -246,6 +529,213 @@ index 487871b..1496eed 100644
  if (toupper(argv[1][0]) == 'P')
      {
      PERIODIC = 1;
+@@ -123,7 +179,7 @@ if (argc != 9 && argc !=11)
+     printf( COMMAND_LINE );
+     exit(-1);
+     }
+-    
++
+ HIST_MIN = atof(argv[1]);
+ HIST_MAX = atof(argv[2]);
+ NUM_BINS = atoi(argv[3]);
+@@ -213,13 +269,13 @@ if (!ave_pdf2)
+     printf("couldn't allocate space for ave_pdf2: %s\n", strerror(errno));
+     exit(errno);
+     }
+- 
++
+ free_ene = (double *) malloc(sizeof(double) * NUM_BINS);
+ if (!free_ene)
+     {
+     printf("couldn't allocate space for free_ene: %s\n", strerror(errno));
+     exit(errno);
+-    }  
++    }
+ 
+ i = get_numwindows(METAFILE);
+ printf("#Number of windows = %d\n", i);
+@@ -248,7 +304,7 @@ assert(i == hist_group->num_windows);
+ // Figure out if we have trajectories at different temperatures.
+ // Missing temperatures are set to -1 in read_metadata, and
+ // since we require that either all trajectories specify a temperature
+-// or all trajectories are assumed to be at the wham temperature, we only 
++// or all trajectories are assumed to be at the wham temperature, we only
+ // have to check one of them
+ if (hist_group->kT[0] > 0)
+     {
+@@ -257,7 +313,7 @@ if (hist_group->kT[0] > 0)
+ else
+     {
+     have_energy = 0;
+-    for (i=0; i< hist_group->num_windows; i++) 
++    for (i=0; i< hist_group->num_windows; i++)
+         {
+         hist_group->kT[i] = kT;
+         }
+@@ -269,7 +325,7 @@ if (!final_f)
+     {
+     printf("couldn't allocate space for final_f: %s\n", strerror(errno));
+     exit(errno);
+-    } 
++    }
+ 
+ free(HISTOGRAM);
+ 
+@@ -305,7 +361,8 @@ while (! is_converged(hist_group) || first )
+         for (i=0; i< NUM_BINS; i++)
+             {
+             coor = calc_coor(i);
+-            printf("%f\t%f\t%f\n", coor, free_ene[i], prob[i]);
++            if (prob[i] != 0.0)
++                printf("%f\t%f\t%f\n", coor, free_ene[i], prob[i]);
+             }
+         printf("\n");
+ 
+@@ -319,7 +376,7 @@ while (! is_converged(hist_group) || first )
+             }
+         }
+     // Cheesy bailout if we're going on too long
+-    if (iteration >= max_iteration) 
++    if (iteration >= max_iteration)
+         {
+         printf("Too many iterations: %d\n", iteration);
+         break;
+@@ -383,11 +440,11 @@ for (i=0; i< num_mc_runs; i++)
+         //printf("Faking %d: %d  %d\n", i,j,hp->num_points);
+         num_used = hp->last - hp->first + 1;
+         mk_new_hist(hp->cum, hp->data, num_used, hp->num_mc_samples, &idum);
+-      
++
+         hist_group->F_old[j] = 0.0;
+         hist_group->F[j] = 0.0;
+         }
+-      
++
+       // perform WHAM iterations on the fake data sets
+       iteration = 0;
+       first = 1;
+@@ -403,7 +460,7 @@ for (i=0; i< num_mc_runs; i++)
+           printf("Too many iterations: %d\n", iteration);
+           break;
+           }
+-        }  
++        }
+       printf("#MC trial %d: %d iterations\n", i, iteration);
+       printf("#PMF values\n");
+       // accumulate the average and stdev of the resulting probabilities
+@@ -419,18 +476,19 @@ for (i=0; i< num_mc_runs; i++)
+       for (j=0; j < NUM_BINS; j++)
+           {
+           pdf = -kT*log(prob[j]);
+-      
++
+           ave_p[j] += prob[j];
+           ave_pdf[j] += pdf;
+           ave_p2[j] += prob[j] * prob[j];
+           ave_pdf2[j] += pdf*pdf;
+           }
+-      for (j=0; j<hist_group->num_windows;j++) 
++      for (j=0; j<hist_group->num_windows;j++)
+           {
+           ave_F[j] += hist_group->F[j] - hist_group->F[0];
+-          ave_F2[j] += hist_group->F[j]*hist_group->F[j] ; 
++          ave_F2[j] += hist_group->F[j]*hist_group->F[j] ;
+           }
+-    } 
++    }
++ if (num_mc_runs == 0) num_mc_runs = 1;
+  for (i=0; i < NUM_BINS; i++)
+    {
+    ave_p[i] /= (double)num_mc_runs;
+@@ -457,12 +515,12 @@ if (!FREEFILE)
+     for (i=0; i< NUM_BINS; i++)
+         {
+         coor = calc_coor(i);
+-        printf("%f\t%f\t%f\t%f\t%f\n", coor, free_ene[i], ave_pdf2[i], 
+-                    prob[i], ave_p2[i]);
++        if (prob[i] != 0.0)
++            printf("%f\t%f\t%f\t%f\t%f\n", coor, free_ene[i], ave_pdf2[i], prob[i], ave_p2[i]);
+         }
+     for (i=0; i<hist_group->num_windows; i++)
+         {
+-        fprintf(FREEFILE,"%d\t%f\t%f\n", i, final_f[i],ave_F2[i]);  
++            printf("%d\t%f\t%f\n", i, final_f[i],ave_F2[i]);
+         }
+ 
+     exit(errno);
+@@ -470,38 +528,37 @@ if (!FREEFILE)
+ else
+     {
+     // write out header
+-    fprintf(FREEFILE, "#Coor\t\tFree\t+/-\t\tProb\t\t+/-\n");
++    fprintf(FREEFILE, "#Coor\t\tFree\t\t+/-\t\tProb\t\t+/-\n");
+     // write out the leading padded values
+     for (i=-numpad; i<0; i++)
+         {
+         coor = calc_coor(i);
+-        fprintf(FREEFILE,"%f\t%f\t%f\t%f\t%f\n", coor, free_ene[NUM_BINS+i], 
+-        ave_pdf2[NUM_BINS+i], 
+-        final_prob[NUM_BINS+i],
+-        ave_p2[NUM_BINS+i]);
++        if (prob[NUM_BINS+1] != 0.0)
++            fprintf(FREEFILE,"%f\t%f\t%f\t%f\t%f\n", coor, free_ene[NUM_BINS+i],
++                    ave_pdf2[NUM_BINS+i], final_prob[NUM_BINS+i], ave_p2[NUM_BINS+i]);
+         }
+     // write out the center values
+     for (i=0; i<NUM_BINS; i++)
+         {
+         coor = calc_coor(i);
+-        fprintf(FREEFILE,"%f\t%f\t%f\t%f\t%f\n", coor, free_ene[i],
+-        ave_pdf2[i],final_prob[i], 
+-        ave_p2[i]);  
++        if (prob[i] != 0.0)
++            fprintf(FREEFILE,"%f\t%f\t%f\t%f\t%f\n", coor, free_ene[i],
++                    ave_pdf2[i],final_prob[i],ave_p2[i]);
+         }
+ 
+     // write out the trailing padded values
+     for (i=0; i<numpad; i++)
+         {
+         coor = calc_coor(NUM_BINS+i);
+-        fprintf(FREEFILE,"%f\t%f\t%f\t%f\t%f\n", coor, free_ene[i], 
+-        ave_pdf2[i],final_prob[i], 
+-        ave_p2[i]); 
++        if (prob[i] != 0.0)
++            fprintf(FREEFILE,"%f\t%f\t%f\t%f\t%f\n", coor, free_ene[i],
++                    ave_pdf2[i],final_prob[i], ave_p2[i]);
+         }
+ 
+-    fprintf(FREEFILE, "#Window\t\tFree\t+/-\t\n");
++    fprintf(FREEFILE, "#Window\t\tFree\t\t+/-\t\n");
+     for (i=0; i<hist_group->num_windows; i++)
+         {
+-        fprintf(FREEFILE,"#%d\t%f\t%f\n", i, final_f[i],ave_F2[i]);  
++        fprintf(FREEFILE,"#%d\t\t%f\t%f\n", i, final_f[i],ave_F2[i]);
+         }
+     }
+ 
+@@ -515,7 +572,7 @@ exit(0);
+ /*
+  *  Perform a single WHAM iteration
+  */
+-void wham_iteration(struct hist_group* hist_group, double *prob, 
++void wham_iteration(struct hist_group* hist_group, double *prob,
+                    int have_energy)
+ {
+ int i,j;
+@@ -535,9 +592,9 @@ for (i=0; i<NUM_BINS; i++)
+         num += (double) get_histval( &(hist_group->hists[j]),i);
+         bias = calc_bias(hist_group,j,coor);
+         bf = exp((hist_group->F_old[j] - bias) / hist_group->kT[j]);
+-        /* If we have energies, then the histograms contain boltzmann 
+-         * factors.  If not, they contain integer counts.  Accordingly, 
+-         * we either use a partition function to normalize, or simply the 
++        /* If we have energies, then the histograms contain boltzmann
++         * factors.  If not, they contain integer counts.  Accordingly,
++         * we either use a partition function to normalize, or simply the
+          * number of points.
+          */
+         if (have_energy)
 diff --git a/wham/wham.h b/wham/wham.h
 index aacc1e8..7d509f2 100644
 --- a/wham/wham.h

tools/swig/lammps.i

@@ -141,6 +141,8 @@ extern int    lammps_extract_variable_datatype(void *handle, const char *name);
 extern int    lammps_set_variable(void *, const char *, const char *);
 extern int    lammps_set_string_variable(void *, const char *, const char *);
 extern int    lammps_set_internal_variable(void *, const char *, double);
+extern int    lammps_variable_info(void *handle, int idx, char *buf, int bufsize);
+extern double lammps_eval(void *handle, const char *expr);
 
 extern void   lammps_gather_atoms(void *, char *, int, int, void *);
 extern void   lammps_gather_atoms_concat(void *, char *, int, int, void *);
@@ -332,6 +334,8 @@ extern int    lammps_extract_variable_datatype(void *handle, const char *name);
 extern int    lammps_set_variable(void *, const char *, const char *);
 extern int    lammps_set_string_variable(void *, const char *, const char *);
 extern int    lammps_set_internal_variable(void *, const char *, double);
+extern int    lammps_variable_info(void *handle, int idx, char *buf, int bufsize);
+extern double lammps_eval(void *handle, const char *expr);
 
 extern void   lammps_gather_atoms(void *, char *, int, int, void *);
 extern void   lammps_gather_atoms_concat(void *, char *, int, int, void *);

unittest/c-library/test_library_objects.cpp

@@ -233,3 +233,34 @@ TEST_F(LibraryObjects, expand)
     EXPECT_THAT((char *)ptr, StrEq("'xx_$(4+5)_$(PI) ${one}-${two}-${three}'"));
     lammps_free(ptr);
 }
+
+TEST_F(LibraryObjects, eval)
+{
+    lammps_get_last_error_message(lmp, nullptr, 1);
+    ::testing::internal::CaptureStdout();
+    lammps_commands_string(lmp, "region box1 block 0 10 0 5 -0.5 0.5\n"
+                                "lattice fcc 0.8\n"
+                                "create_box 1 box1\n"
+                                "create_atoms 1 box\n"
+                                "mass * 1.0\n"
+                                "pair_style lj/cut 4.0\n"
+                                "pair_coeff * * 1.0 1.0\n"
+                                "variable t equal 15.0\n"
+                                "velocity all create 1.5 532656\n"
+                                "fix 1 all nve\n"
+                                "run 0 post no\n");
+    lammps_command(lmp, "variable one    index     1 2 3 4");
+    lammps_command(lmp, "variable two    equal     2");
+    std::string output = ::testing::internal::GetCapturedStdout();
+    if (verbose) std::cout << output;
+    ASSERT_EQ(lammps_has_error(lmp), 0);
+
+    EXPECT_DOUBLE_EQ(lammps_eval(lmp, "4+5"), 9.0);
+    EXPECT_EQ(lammps_has_error(lmp), 0);
+    EXPECT_DOUBLE_EQ(lammps_eval(lmp, "v_one / 2.0"), 0.5);
+    EXPECT_EQ(lammps_has_error(lmp), 0);
+    EXPECT_DOUBLE_EQ(lammps_eval(lmp, "count(all)"), 36.0);
+    EXPECT_EQ(lammps_has_error(lmp), 0);
+    EXPECT_DOUBLE_EQ(lammps_eval(lmp, "pe"), -3.9848867644689534);
+    EXPECT_EQ(lammps_has_error(lmp), 0);
+}

unittest/python/python-commands.py

@@ -538,6 +538,27 @@ create_atoms 1 single &
         expanded = self.lmp.expand("'xx_$(4+5)_$(PI) ${one}-${two}-${three}'")
         self.assertEqual(expanded, "'xx_$(4+5)_$(PI) ${one}-${two}-${three}'")
 
+    def test_eval(self):
+        self.lmp.commands_string(
+            """region box1 block 0 10 0 5 -0.5 0.5
+            lattice fcc 0.8
+            create_box 1 box1
+            create_atoms 1 box
+            mass * 1.0
+            pair_style lj/cut 4.0
+            pair_coeff * * 1.0 1.0
+            variable t equal 15.0
+            velocity all create 1.5 532656
+            fix 1 all nve
+            run 0 post no""")
+        self.lmp.command("variable one    index     1 2 3 4")
+        self.lmp.command("variable two    equal     2")
+
+        self.assertEqual(self.lmp.eval("4+5"), 9.0)
+        self.assertEqual(self.lmp.eval("v_one / 2.0"), 0.5)
+        self.assertEqual(self.lmp.eval("count(all)"), 36.0)
+        self.assertEqual(self.lmp.eval("pe"), -3.9848867644689534)
+
     def test_get_thermo(self):
         self.lmp.command("units lj")
         self.lmp.command("atom_style atomic")