I wrote Hartree to help my wife with her processing of Gaussian 09 log files.
The project uses Maven as its build tool and
has cmayes-common as a dependency.
Since cmayes-common is not yet in Maven Central, you will need to build and install
this library locally.
$ java -version
Hartree requires Java 8 or better. If the above command doesn't show 8.x or above, you will need to upgrade your Java version. If you're able to, I'd suggest using SDKMan for managing local JDK versions.
Maven is a Java library and can be unpacked in a local directory and added to your classpath. For example:
$ cd ~/apps
$ wget http://www-eu.apache.org/dist/maven/maven-3/3.6.3/binaries/apache-maven-3.6.3-bin.tar.gz
$ tar zxvf apache-maven-3.6.3-bin.tar.gz
$ ln -s apache-maven-3.6.3 maven
$ echo 'export PATH=$PATH:$HOME/apps/maven/bin' >> ~/.bashrc
$ . ~/.bashrc
$ mvn --version
The last command should produce an output starting with something similar to Apache Maven 3.6.3.
This library also uses Maven, so be sure that it's installed as per the previous section.
$ cd ~/code
$ git clone https://github.com/cmayes/cmayes-common
$ cd cmayes-common
$ git checkout cmayes-common-1.1.7
$ mvn clean install
This will pull down the library's dependencies, run the build for the 1.1.7 release, and install
the release's JAR in your local Maven repository (usually ~/.m2/repository).
Once you have Maven and cmayes-common installed, you can build Hartree.
$ cd ~/code
$ git clone https://github.com/team-mayes/hartree
$ cd hartree
$ git checkout hartree-1.3.2
$ mvn clean package
Note that you should check out the tag that corresponds to the release that you wish to build.
The artifact at hartree-cli/target/hartree-cli-1.3.2.jar (or whichever version you've built)
is what you'll use.
The basic usage is:
$ java -jar hartree-cli-1.3.2.jar
Argument not one of (norm,snap,cpsnap,test,therm,lowen)
java org.cmayes.hartree.Main [options...] (norm,snap,cpsnap,test,therm,lowen)
Available arguments:
norm : Handles normal mode evaluation
snap : Provides a snapshot of calculation data
cpsnap : Provides a snapshot of calculation data including Cremer-Pople coordinates
test : Test mode (no handling performed)
therm : Handles thermo calculations
lowen : creates Gaussian input files for the lowest energy in the input files
Available options:
--cfgfile (-c) CFG : Configuration settings file (required for DB inserts)
--directory (-d) INDIR : The base directory of the files to process
--extensions (-e) EXTS : Extensions to include in input directory searches
(.log and .out by default)
--file (-f) INFILE : The file to process
--ion (-i) ION : The ion element type to use.
--mediatype (-m) MEDIA : The media type to use instead of the default.
--outdir (-o) OUTDIR : The output directory for result files
--proctype (-p) PROC : The processor type to use instead of the default.
--projname (-n) PROJ : The name of this data's project (required for DB
inserts)
--tags (-t) TAGS : Categories that describe the input data
This tells Java to run the main class in the JAR. The output is a usage message because we have not specified an operation.
Let's try the normal mode option:
$ java -jar hartree-cli-1.3.2.jar norm
No input file or directory specified.
java org.cmayes.hartree.Main [options...] (norm,snap,cpsnap,test,therm,lowen)
Available arguments:
norm : Handles normal mode evaluation
snap : Provides a snapshot of calculation data
cpsnap : Provides a snapshot of calculation data including Cremer-Pople coordinates
test : Test mode (no handling performed)
therm : Handles thermo calculations
lowen : creates Gaussian input files for the lowest energy in the input files
Available options:
--cfgfile (-c) CFG : Configuration settings file (required for DB inserts)
--directory (-d) INDIR : The base directory of the files to process
--extensions (-e) EXTS : Extensions to include in input directory searches
(.log and .out by default)
--file (-f) INFILE : The file to process
--ion (-i) ION : The ion element type to use.
--mediatype (-m) MEDIA : The media type to use instead of the default.
--outdir (-o) OUTDIR : The output directory for result files
--proctype (-p) PROC : The processor type to use instead of the default.
--projname (-n) PROJ : The name of this data's project (required for DB
inserts)
--tags (-t) TAGS : Categories that describe the input data
As the error message indicate, We need to specify a source to process. Let's do a single file first.
$ java -jar hartree-cli-1.3.2.jar norm -f glucose5m062xEtOHnorm.log
Normal mode summary for file glucose5m062xEtOHnorm.log
Highest DoF percentages by dihedral:
...
The report is dumped to standard out. We could redirect this to a file, but it's just as easy to point Hartree to an output directory:
$ java -jar hartree-cli-1.3.2.jar norm -f glucose5m062xEtOHnorm.log -o out
05:19:29.377 [main] DEBUG org.cmayes.hartree.Main - Created out dir
/home/cmayes/Downloads/out
This does not produce any output aside from a debug message about creating the out dir (it won't write that message if the dir exists). In the output dir, we have one file:
$ ls out
glucose5m062xEtOHnorm.log-norm.txt
This is the normal mode report. It's named for the log file with "-norm.txt" at the end.
Now, let's try processing a whole directory at once:
$ java -jar hartree-cli-1.3.2.jar norm -d ~/g09/ -o out
line 741:11 mismatched input '45' expecting CPUTAG
...
This generates a lot of noise because most of my test files do not have normal mode data. The result in the out dir:
$ ls out
GL_THF_rev4.log-norm.txt
glucose5m062xEtOHnorm.log-norm.txt
init_gas_rad2_3.out-norm.txt
init_THF+negFAIL.log-norm.txt
init_water+pos4.log-norm.txt
init2thfTS+camAgain3.log-norm.txt
levoglucosan8spTooLittleData.log-norm.txt
m-glucose_gasm062x.out-norm.txt
m-glucose_waterm062xspTooLittleData.out-norm.txt
methanolmp4mp2TooLittleData.log-norm.txt
It's processed all of the input files in the same way, but since most of the input files didn't have any normal mode data, they look like this:
$ cat out/GL_THF_rev4.log-norm.txt
Normal mode summary for file GL_THF_rev4.log
No normal mode data found.
It should handle nested directories such that input files in sub-directories are written to the same sub-directory in the output directory.
The goal of this program is to summarize output from a Gaussian run with the freq=internalmodes.
For each normal mode found, it will sum the Relative Weights (%) from Angle Bending (A), Bond Stretching (R), and Dihedral Angles (D). For dihedral angles, it further added together all the separate dihedral angles that share the same two central atoms (here called a "dihedral pair"). For example, the 64 lines of Gaussian output for Normal Mode 1 in hartree-antlr/target/test-classes/files/g09/glucose5m062xEtOHnorm.log, which includes:
----------------------------
! Normal Mode 1 !
-------------------------- --------------------------
! Name Definition Value Relative Weight (%) !
--------------------------------------------------------------------------------
! A19 A(3,4,5) 0.0243 0.6 !
! D1 D(7,1,2,3) 0.0681 1.8 !
! D2 D(7,1,2,8) 0.0686 1.8 !
! D3 D(7,1,2,14) 0.0735 2.0 !
! D4 D(11,1,2,3) 0.0726 1.9 !
...
! D66 D(24,6,12,22) -0.0279 0.7 !
--------------------------------------------------------------------------------
The dihedral angles D1 through D9 all describe rotation around the 1-2 bond (pair) and thus are summed in the hartree output file. The hartree output for Normal Mode 1 summarizes the data in 14 lines as:
=== Normal mode 1 ===
Frequency : 75.23
Angle Bending (A): 0.60
Bond Stretching (R): 0.00
Dihedral pairs (D):
( 1, 2): 17.60
( 2, 3): 11.00
( 2, 8): 5.00
( 3, 4): 9.30
( 4, 5): 25.20
( 4, 10): 4.40
( 5, 6): 9.60
( 5, 11): 6.40
( 6, 12): 2.20
Note that total sum of the original relative weights in the Guassian output file was 91.3%. Thus, the total relative weights in the new summary also do not equal 100%.
This "snapshot" option summarizes salient output from the many-line Guassian output file(s) into one line per file. There may be multiple listings of some properties (i.e. electronic energy). The listed value will be the last one in the file. The column headings in the resulting csv file are:
- File Name: repeats the file name where the data was obtained
- Solvent type: if an implicit solvent was specified (i.e. water) it will be noted here. Otherwise, the program will list "N/A"
- Stoichiometry: the molecular formula for the system, i.e. C5H10O
- Charge: total system charge
- Mult: system multiplicity
- Functional: functional used. Note that this will be repeated as listed in the output, not in the input, so you may see "RM062X" instead of "m062x"
- Basis Set: basis set, if applicable ("N/A" if semiempirical)
- Energy (A.U.): Last-listed electronic energy
- dipole: total dipole moment for the system
- ZPE (Hartrees): the zero point energy; "N/A" if no calculated (no frequency calculated)
- G298 (Hartrees): Gibbs free energy at 298 K (if calculated)
- H298 (Hartrees): Enthalpy at 298 K (if calculated)
- Freq 1: Lowest frequency (wavenumber) in cm^-1; since Gaussian sorts frequencies, listing the lowest first, and since it stores imaginary numbers as negative numbers, you can use this to determine if the structure is a local minimum (if so, this will be a positive number)
- Freq 2: Second-lowest frequency (wavenumber) in cm^-1; if Freq 1 is listed as a negative number and Freq 2 is positive, then the structure is a saddle point (transition state)
This option produces one-line summaries of each output file (as in the "snap" option). It also provides the following columns. Some were created for very specific tasks and thus may not be generally useful.
The "CP" of cpsnap stands for Cremer-Pople. While CP parameters are generally defined for many ring sizes, this program assumes only 6-membered rings.
- phi: Cremer-Pople (CP) puckering parameter phi, calculated
- theta: CP theta
- Q: CP Q
- Pucker: the closest IUPAC recognized puckering designation, as calculated based on arc length
- HM1 (deg): HM stands for "hydroxymethyl" and this is the O5-C5-C6-O6 dihedral angle; these calcs were used to determine the orientation of the hydroxymethyl arm on sugars such as glucose, GlcNAC, and mannose. "N/A" is returned for xylose.
- HM2 (deg): as above, with this showing the C4-C5-C6-O6 dihedral angle
- AC1 (deg): valid for GlcNAc (N/A otherwise); angle specifying orientation of acetyl group: C1-C2-N-C
- AC2 (deg): as above, but C3-C2-N-C
- Ano1 (deg): the C5-O5-C1-O1 dihedral angle, used to determine the stereochemistry at the anomeric carbon
- Ano2 (deg): the C3-C2-C1-O1 dihedral angle (the other anomeric carbon angle)
- R1 (A): these are the bond length around the 6-membered ring: O5-C1
- R2 (A): C1-C2
- R3 (A): C2-C3
- R4 (A): C3-C4
- R5 (A): C4-C5
- R6 (A): C5-O5
- O1 (A): For a six-membered ring sugar, the C1-O1 distance. Note that the there is no "O5" output. That is because for the sugars, the "O5" is part of the ring and thus the distances between O5 and C5 (and O5 and C1) are already output above.
- O2 (A): C2-O2 distance (C2-N if GlcNAc)
- O3 (A): C3-O3 distance
- O4 (A): C4-O4 distance
- O6 (A): C6-O6 distance (N/A if the sugar is xylose)
- Ion1 (A): distance between a specified atom type and each of up to 6 oxygen atoms (O1 to O6) (assumes one such atom type, as created for when I looked at ions like Ca or Na interacting with a glucose molecule). By default, it will look for a sodium ion (Na or 11). This can be changes with the -i or --ion option. Note that the whole element name must be used (i.e. "-i calcium").
- Ion2 (A): ion-O2
- Ion3 (A): ion-O3
- Ion4 (A): ion-O4
- Ion5 (A): ion-O5
- Ion6 (A): ion-O6
Test function; no output created
Not sure that this function is complete.
Created to help restart a calculation from the lowest energy point (determined from looking at "SCF: Done"), which may not be the last point. Sometimes, an optimization goes off course and restarting from an intermediate position is helpful. The output (xyz coordinates) is added to a template. An example template can be found at /hartree-common/src/test/resources/files/tpl/lowtpl