first commit

.gitignore

@@ -0,0 +1,8 @@
+# Ignore everything
+*
+
+# But track these files
+!.gitignore
+!dump2rotorang.py
+!example.lammpstrj
+!Readme.org

Readme.org

@@ -0,0 +1,188 @@
+#+TITLE: dump2rotorang.py
+#+AUTHOR:
+#+DATE:
+
+* Overview
+
+=dump2rotorang.py= analyzes LAMMPS trajectory files and calculates
+orientation and z-position distributions of benzene molecules.
+
+The script reads LAMMPS dump trajectory files =*.lammpstrj=, extracts the
+coordinates of selected atoms in each molecule, and evaluates:
+
+- the in-plane orientation angle of each benzene molecule
+- the center z-position of each benzene molecule
+
+The resulting histograms are saved as Python pickle files.
+
+* Requirements
+
+This script requires Python 3 and the following Python packages:
+
+- numpy
+- pandas
+
+Install them, for example, using pip:
+
+#+begin_src sh
+pip install numpy pandas
+#+end_src
+
+* Input files
+
+The script searches for trajectory files with the following filename
+patterns:
+
+#+begin_src text
+D3d-o_*K.lammpstrj
+D3d-p_*K.lammpstrj
+#+end_src
+
+For example:
+
+#+begin_src text
+D3d-o_300K.lammpstrj
+D3d-o_400K.lammpstrj
+D3d-p_300K.lammpstrj
+D3d-p_400K.lammpstrj
+#+end_src
+
+The temperature is extracted from the filename using the pattern:
+
+#+begin_src text
+_(number)K.lammpstrj
+#+end_src
+
+For example, =D3d-o_300K.lammpstrj= is interpreted as 300 K.
+
+* LAMMPS dump format
+
+The trajectory file is assumed to be a LAMMPS dump file containing
+unwrapped atomic coordinates:
+
+#+begin_src text
+xu yu zu
+#+end_src
+
+The script also assumes that the dump file contains a =mol= column, which
+is used to identify molecules.
+
+A typical header should include columns such as:
+
+#+begin_src text
+ITEM: ATOMS id mol type xu yu zu
+#+end_src
+
+* Analysis details
+
+The benzene atoms are selected by the following atom indices within each
+molecule:
+
+#+begin_src python
+benzene_id = [24, 25, 26, 27, 28, 29]
+#+end_src
+
+These indices are hard-coded in the script.
+
+The orientation of each benzene molecule is calculated from the vector
+between:
+
+- the center of the selected benzene atoms
+- the midpoint of atoms 25 and 26 in the selected benzene atom list
+
+The orientation angle is calculated using:
+
+#+begin_src python
+np.arctan2(vec[:, 1], vec[:, 0])
+#+end_src
+
+The z-position is calculated as the mean z-coordinate of the selected
+benzene atoms.
+
+* Output files
+
+The script creates the following pickle files:
+
+#+begin_src text
+D3d-o.pkl
+D3d-p.pkl
+#+end_src
+
+Each pickle file contains a dictionary indexed by temperature.
+
+For example:
+
+#+begin_src python
+dset[300]["orientation"]
+#+end_src
+
+contains a pandas DataFrame with the following columns:
+
+| Column    | Description                              |
+|-----------+------------------------------------------|
+| angle     | Bin center of orientation angle           |
+| angle_y   | Histogram count of orientation angle      |
+| zpos      | Bin center of z-position                  |
+| zpos_y    | Histogram count of z-position             |
+
+* Usage
+
+Place =dump2rotorang.py= in the directory containing the LAMMPS trajectory
+files, then run:
+
+#+begin_src sh
+python dump2rotorang.py
+#+end_src
+
+If trajectory files matching the expected patterns exist, the script will
+print the processed filenames and create pickle files.
+
+Example:
+
+#+begin_src text
+D3d-o_300K.lammpstrj
+D3d-o_400K.lammpstrj
+D3d-o.pkl was created.
+D3d-p_300K.lammpstrj
+D3d-p_400K.lammpstrj
+D3d-p.pkl was created.
+#+end_src
+
+* Example
+
+If =example.lammpstrj= is included as a sample trajectory file, rename or
+copy it to match the expected naming rule before running the script:
+
+#+begin_src sh
+cp example.lammpstrj D3d-o_300K.lammpstrj
+python dump2rotorang.py
+#+end_src
+
+This will create:
+
+#+begin_src text
+D3d-o.pkl
+#+end_src
+
+* Notes
+
+- The atom indices used for benzene are currently hard-coded.
+- The script assumes that all molecules have the same number of atoms.
+- The orientation histogram is calculated in the range from =-pi= to =pi=.
+- The z-position histogram is calculated in the range from =24= to =25=.
+- If the target system or molecule definition is changed, the values of
+  =benzene_id= and the histogram range for z-position may need to be
+  modified.
+
+* Files managed in this repository
+
+This repository is intended to manage only the following files:
+
+#+begin_src text
+dump2rotorang.py
+example.lammpstrj
+README.org
+.gitignore
+#+end_src
+
+

dump2rotorang.py

@@ -0,0 +1,98 @@
+#!/usr/bin/env python
+import numpy as np
+import io
+import re
+import pickle
+import glob
+import pandas as pd
+
+
+class LammpsTrj():
+    def __init__(self, trjfile):
+        self.fp = open(trjfile)
+        lines = [self.fp.readline() for i in range(4)]
+        self.atoms = int(lines[-1])
+        self.fp.seek(0)
+        self.getOnestep()
+        # print(self.df)
+        self.fp.seek(0)
+        self.molecules = np.unique(self.df["mol"]).shape[0]
+
+    def __del__(self):
+        self.fp.close()
+
+    def getOnestep(self):
+        block = self.atoms + 9
+        lines = [self.fp.readline() for i in range(block)]
+        if lines[0] == "":
+            return 0
+        self.step = int(lines[1])
+        self.lx = float(lines[5].split()[1])
+        self.ly = float(lines[6].split()[1])
+        self.lz = float(lines[7].split()[1])
+        header = lines[8].split()[2:]
+        data = io.StringIO(" ".join(lines[9:]))
+        self.df = pd.read_csv(data, delimiter=r"\s+", header=None)
+        self.df.columns = header
+        return 1
+
+    def getOrientation(self, benzene_id):
+        xyz = self.df[["xu", "yu", "zu"]].to_numpy()
+        xyz = xyz.reshape((self.molecules, -1, 3))
+        benzene_id = np.array(benzene_id, dtype=int) - 1
+        xyz = xyz[:, benzene_id, :]
+        g = xyz.mean(axis=1)
+        p = xyz[:, [1, 2], :].mean(axis=1)
+        vec = p - g
+        angle = np.arctan2(vec[:, 1], vec[:, 0])
+        return angle
+        # for i, xyz_i in enumerate(xyz):
+        #     body = "{}\n\n".format(benzene_id.shape[0])
+        #     for r in xyz_i:
+        #         body += "C {} {} {}\n".format(*r)
+        #     outfile = f"{i:04d}.xyz"
+        #     with open(outfile, "w") as o:
+        #         o.write(body)
+
+    def getZposition(self, benzene_id):
+        xyz = self.df[["xu", "yu", "zu"]].to_numpy()
+        xyz = xyz.reshape((self.molecules, -1, 3))
+        benzene_id = np.array(benzene_id, dtype=int) - 1
+        xyz = xyz[:, benzene_id, :]
+        zpos = xyz.mean(axis=1)[:, 2]
+        return zpos
+
+
+def run(trjfile):
+    benzene_id = [24, 25, 26, 27, 28, 29]
+    lmp = LammpsTrj(trjfile)
+    orientation = np.empty((0, lmp.molecules))
+    zpos = np.empty((0, lmp.molecules))
+    while lmp.getOnestep():
+        # print(lmp.step)
+        orientation_ = lmp.getOrientation(benzene_id)
+        orientation = np.vstack((orientation, orientation_))
+        zpos_ = lmp.getZposition(benzene_id)
+        zpos = np.vstack((zpos, zpos_))
+    hist, x = np.histogram(orientation, bins=np.linspace(-np.pi, np.pi, 200))
+    df = pd.DataFrame()
+    df["angle"] = x[0:-1] + (x[1]-x[0])*0.5
+    df["angle_y"] = hist
+    hist, x = np.histogram(zpos, bins=np.linspace(24, 25, 200))
+    df["zpos"] = x[0:-1] + (x[1]-x[0])*0.5
+    df["zpos_y"] = hist
+    temp = int(re.search(r"_(\d+)K\.", trjfile).group(1))
+    dset[temp] = {}
+    dset[temp]["orientation"] = df
+
+
+for k in ["D3d-o", "D3d-p"]:
+    dset = {}
+    files = sorted(glob.glob(f"{k}_*K.lammpstrj"))
+    for f in files:
+        print(f)
+        run(f)
+    outfile = f"{k}.pkl"
+    with open(outfile, "wb") as f:
+        pickle.dump(dset, f)
+    print(f"{outfile} was created.")