system ener is always double. print code version and info of building in train and md

Author: Han Wang

README.md

@@ -15,7 +15,7 @@
 	    - [Install tensorflow's C++ interface](#install-tensorflows-c-interface)
 	    - [Install xdrfile](#install-xdrfile)
 	    - [Install DeePMD-kit](#install-deepmd-kit)
-	    - [Install LAMMPS's DeePMD-kit module](#install-lammps-deepmd-kit-module)
+	    - [Install LAMMPS's DeePMD-kit module](#install-lammpss-deepmd-kit-module)
 	    - [Build DeePMD-kit with GPU support](#build-deepmd-kit-with-gpu-support)
 - [Use DeePMD-kit](#use-deepmd-kit)
 	- [Prepare data](#prepare-data)
@@ -323,7 +323,11 @@ $ cat force.raw
 ```
 This `force.raw` contains 3 frames with each frame having the forces of 2 atoms, thus it has 3 lines and 6 columns. Each line provides all the 3 force components of 2 atoms in 1 frame. The first three numbers are the 3 force components of the first atom, while the second three numbers are the 3 force components of the second atom. The coordinate file `coord.raw` is organized similarly. In `box.raw`, the 9 components of the box vectors should be provided on each line. In `virial.raw`, the 9 components of the virial tensor should be provided on each line. The number of lines of all raw files should be identical.
 
-We assume that the atom types do not change in all frames. It is provided by `type.raw`, which has one line with the types of atoms written one by one. The atom types should be integers.
+We assume that the atom types do not change in all frames. It is provided by `type.raw`, which has one line with the types of atoms written one by one. The atom types should be integers. For example the `tyep.raw` of a system that has 2 atoms with 0 and 1:
+```bash
+$ cat type.raw
+0 1
+```
 
 The second format is the data sets of `numpy` binary data that are directly used by the training program. User can use the script `$deepmd_source_dir/data/raw/raw_to_set.sh` to convert the prepared raw files to data sets. For example, if we have a raw file that contains 6000 frames, 
 ```bash

data/raw/raw_to_set.sh

@@ -7,14 +7,15 @@ if test $# -ge 1; then
 fi
 
 rm -fr set.*
-echo nframe is `cat energy.raw | wc -l`
+echo nframe is `cat box.raw | wc -l`
 echo nline per set is $nline_per_set
 
 split box.raw	 -l $nline_per_set -d -a 3 box.raw
 split coord.raw	 -l $nline_per_set -d -a 3 coord.raw
 test -f energy.raw && split energy.raw -l $nline_per_set -d -a 3 energy.raw
 test -f force.raw  && split force.raw  -l $nline_per_set -d -a 3 force.raw
 test -f virial.raw && split virial.raw -l $nline_per_set -d -a 3 virial.raw
+test -f atom_ener.raw && split atom_ener.raw -l $nline_per_set -d -a 3 atom_ener.raw
 
 nset=`ls | grep box.raw[0-9] | wc -l`
 nset_1=$(($nset-1))
@@ -30,6 +31,7 @@ do
   test -f energy.raw$pi && mv energy.raw$pi set.$pi/energy.raw
   test -f force.raw$pi  && mv force.raw$pi  set.$pi/force.raw
   test -f virial.raw$pi && mv virial.raw$pi set.$pi/virial.raw
+  test -f atom_ener.raw$pi && mv atom_ener.raw$pi set.$pi/atom_ener.raw
 
   cd set.$pi
   python -c 'import numpy as np; data = np.loadtxt("box.raw"   ); data = data.astype (np.float32); np.save ("box",    data)'
@@ -54,6 +56,13 @@ if os.path.isfile("virial.raw"):
    data = np.loadtxt("virial.raw"); 
    data = data.astype (np.float32); 
    np.save ("virial", data)
+'
+  python -c \
+'import numpy as np; import os.path; 
+if os.path.isfile("atom_ener.raw"): 
+   data = np.loadtxt("atom_ener.raw"); 
+   data = data.astype (np.float32); 
+   np.save ("atom_ener", data)
 '
   rm *.raw
   cd ../

examples/train/water.json

@@ -1,4 +1,5 @@
 {
+    "with_distrib":	false,
     "_comment": " model parameters",
     "use_smooth":	false,
     "sel_a":		[16, 32],
@@ -35,8 +36,8 @@
     "_comment": " frequencies counted in batch",
     "disp_file":	"lcurve.out",
     "disp_freq":	100,
-    "numb_test":	100,
-    "save_freq":	100,
+    "numb_test":	10,
+    "save_freq":	1000,
     "save_ckpt":	"model.ckpt",
     "load_ckpt":	"model.ckpt",
     "disp_training":	true,

examples/train/water_smth.json

@@ -34,8 +34,8 @@
     "_comment": " frequencies counted in batch",
     "disp_file":	"lcurve.out",
     "disp_freq":	100,
-    "numb_test":	50,
-    "save_freq":	100,
+    "numb_test":	10,
+    "save_freq":	1000,
     "save_ckpt":	"model.ckpt",
     "load_ckpt":	"model.ckpt",
     "disp_training":	true,

source/CMakeLists.txt

@@ -1,9 +1,32 @@
 cmake_minimum_required(VERSION 3.0)
 project(DeePMD)
 
+find_package(Git)
+if(GIT_FOUND)
+  execute_process(
+    COMMAND git log -1 --format=%h
+    WORKING_DIRECTORY ${CMAKE_SOURCE_DIR}
+    OUTPUT_VARIABLE GIT_HASH
+    OUTPUT_STRIP_TRAILING_WHITESPACE
+    )
+  execute_process(
+    COMMAND git rev-parse --abbrev-ref HEAD
+    WORKING_DIRECTORY ${CMAKE_SOURCE_DIR}
+    OUTPUT_VARIABLE GIT_BRANCH
+    OUTPUT_STRIP_TRAILING_WHITESPACE
+    )
+  execute_process(
+    COMMAND git show -s --format=%ci ${GIT_HASH}
+    WORKING_DIRECTORY ${CMAKE_SOURCE_DIR}
+    OUTPUT_VARIABLE GIT_DATE
+    OUTPUT_STRIP_TRAILING_WHITESPACE
+    )
+endif(GIT_FOUND)
+
 # set(CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} -O0")
 # set(CMAKE_C_FLAGS_DEBUG "${CMAKE_C_FLAGS_DEBUG} -O0")
 
+# global defines
 list (APPEND CMAKE_MODULE_PATH ${PROJECT_SOURCE_DIR}/cmake/)
 list (APPEND CMAKE_CXX_FLAGS "-std=c++11 -Wno-ignored-attributes")
 
@@ -61,6 +84,7 @@ set (LIB_DEEPMD_OP	"deepmd_op")
 set (LIB_DEEPMD_NATIVE	"deepmd_native_md")
 set (LIB_DEEPMD_IPI	"deepmd_ipi")
 
+include_directories(${CMAKE_BINARY_DIR}/lib/)
 add_subdirectory (lib/)
 add_subdirectory (op/)
 add_subdirectory (train/)

source/cmake/cmake_lammps.cmake.in

@@ -15,3 +15,9 @@ file (
   TYPE FILE 
   FILES "@CMAKE_BINARY_DIR@/lmp/env.sh"
 )
+
+file (
+  INSTALL DESTINATION "${LMP_INSTALL_PREFIX}" 
+  TYPE FILE 
+  FILES "@CMAKE_BINARY_DIR@/lmp/pair_nnp.h"
+)

source/ipi/driver.cc

@@ -192,7 +192,7 @@ int main(int argc, char * argv[])
       for (unsigned dd = 0; dd < dbox.size(); ++dd) dbox_[dd] = dbox[dd];
       vector<float> dforce_(dforce.size(), 0);
       vector<float> dvirial_(dvirial.size(), 0);
-      float dener_ = 0;
+      double dener_ = 0;
       nnp_inter.compute (dener_, dforce_, dvirial_, dcoord_, dtype, dbox_);   
       for (unsigned dd = 0; dd < dforce.size(); ++dd) dforce_tmp[dd] = dforce_[dd];	
       for (unsigned dd = 0; dd < dvirial.size(); ++dd) dvirial[dd] = dvirial_[dd];	

source/lib/CMakeLists.txt

@@ -2,9 +2,14 @@
 
 set (libname ${LIB_DEEPMD})
 
+configure_file(
+  ${CMAKE_CURRENT_SOURCE_DIR}/include/version.h.in
+  version.h
+  @ONLY
+)
+
 file(GLOB LIB_SRC src/*.cc src/*.cpp)
-file(GLOB INC_SRC include/*.h)
-# include_directories ("include/")
+file(GLOB INC_SRC include/*.h ${CMAKE_CURRENT_BINARY_DIR}/version.h)
 
 add_library(${libname} SHARED ${LIB_SRC})
 

source/lib/include/NNPInter.h

@@ -7,14 +7,17 @@
 #include "tensorflow/core/framework/shape_inference.h"
 
 #include <vector>
+#include "version.h"
 
 using namespace tensorflow;
 using namespace std;
 
 #ifdef HIGH_PREC
 typedef double VALUETYPE;
+typedef double ENERGYTYPE;
 #else 
 typedef float  VALUETYPE;
+typedef double ENERGYTYPE;
 #endif
 
 struct LammpsNeighborList 
@@ -52,31 +55,32 @@ class NNPInter
   NNPInter () ;
   NNPInter  (const string & model);
   void init (const string & model);
+  void print_summary(const string &pre) const;
 public:
-  void compute (VALUETYPE &			ener,
+  void compute (ENERGYTYPE &			ener,
 		vector<VALUETYPE> &		force,
 		vector<VALUETYPE> &		virial,
 		const vector<VALUETYPE> &	coord,
 		const vector<int> &		atype,
 		const vector<VALUETYPE> &	box, 
 		const int			nghost = 0);
-  void compute (VALUETYPE &			ener,
+  void compute (ENERGYTYPE &			ener,
 		vector<VALUETYPE> &		force,
 		vector<VALUETYPE> &		virial,
 		const vector<VALUETYPE> &	coord,
 		const vector<int> &		atype,
 		const vector<VALUETYPE> &	box, 
 		const int			nghost,
 		const LammpsNeighborList &	lmp_list);
-  void compute (VALUETYPE &			ener,
+  void compute (ENERGYTYPE &			ener,
 		vector<VALUETYPE> &		force,
 		vector<VALUETYPE> &		virial,
 		vector<VALUETYPE> &		atom_energy,
 		vector<VALUETYPE> &		atom_virial,
 		const vector<VALUETYPE> &	coord,
 		const vector<int> &		atype,
 		const vector<VALUETYPE> &	box);
-  void compute (VALUETYPE &			ener,
+  void compute (ENERGYTYPE &			ener,
 		vector<VALUETYPE> &		force,
 		vector<VALUETYPE> &		virial,
 		vector<VALUETYPE> &		atom_energy,
@@ -105,22 +109,22 @@ class NNPInterModelDevi
   NNPInterModelDevi  (const vector<string> & models);
   void init (const vector<string> & models);
 public:
-  void compute (VALUETYPE &			ener,
+  void compute (ENERGYTYPE &			ener,
   		vector<VALUETYPE> &		force,
   		vector<VALUETYPE> &		virial,
   		vector<VALUETYPE> &		model_devi,
   		const vector<VALUETYPE> &	coord,
   		const vector<int> &		atype,
   		const vector<VALUETYPE> &	box);
-  void compute (vector<VALUETYPE> &		all_ener,
+  void compute (vector<ENERGYTYPE> &		all_ener,
 		vector<vector<VALUETYPE> > &	all_force,
 		vector<vector<VALUETYPE> > &	all_virial,
 		const vector<VALUETYPE> &	coord,
 		const vector<int> &		atype,
 		const vector<VALUETYPE> &	box,
 		const int			nghost,
 		const LammpsNeighborList &	lmp_list);
-  void compute (vector<VALUETYPE> &		all_ener,
+  void compute (vector<ENERGYTYPE> &		all_ener,
 		vector<vector<VALUETYPE> > &	all_force,
 		vector<vector<VALUETYPE> > &	all_virial,
 		vector<vector<VALUETYPE> > &	all_atom_energy,
@@ -131,13 +135,14 @@ class NNPInterModelDevi
 		const int			nghost,
 		const LammpsNeighborList &	lmp_list);
   VALUETYPE cutoff () const {return rcut;};
+#ifndef HIGH_PREC
+  void compute_avg (ENERGYTYPE &		dener,
+		    const vector<ENERGYTYPE > &	all_energy);
+#endif
   void compute_avg (VALUETYPE &			dener,
 		    const vector<VALUETYPE > &	all_energy);
-  void compute_avg (vector<VALUETYPE> &			avg,
+  void compute_avg (vector<VALUETYPE> &		avg,
 		    const vector<vector<VALUETYPE> > &	xx);
-  void compute_std   (VALUETYPE &			std,
-		      const VALUETYPE &			avg,
-		      const vector<VALUETYPE >&		xx);
   void compute_std_e (vector<VALUETYPE> &		std,
 		      const vector<VALUETYPE> &		avg,
 		      const vector<vector<VALUETYPE> >&	xx);

source/lib/include/version.h.in

@@ -0,0 +1,17 @@
+#pragma once
+
+#include <string>
+using namespace std;
+
+#ifdef HIGH_PREC
+const string global_float_prec="double";
+#else 
+const string global_float_prec="float";
+#endif
+
+const string global_install_prefix="@CMAKE_INSTALL_PREFIX@";
+const string global_git_hash="@GIT_HASH@";
+const string global_git_date="@GIT_DATE@";
+const string global_git_branch="@GIT_BRANCH@";
+const string global_tf_include_dir="@TensorFlow_INCLUDE_DIRS@";
+const string global_tf_lib_dir="@TensorFlow_LIBRARY_PATH@";