thermotar

For importing LAMMPS log and chunk files, with a light bit of analysing too.

Docs are located at: chappertron.github.io/thermotar/

Installation:

Quick install: Clone the repository and run from the root directory of the project:

pip install .

Alternatively build without explicitly cloning first:

pip install git+https://github.com/chappertron/thermotar

Currently no pip or conda distribution.

Basic usage

Reading a lammps log file:

  import thermotar as th
  # Read LAMMPS log file
  # By default reads the last runs output
  thermo = th.create_thermos('log.lammps')

  import matplotlib.pyplot as plt
  plt.plot(thermo.Step,thermo.Temp)
  plt.show()

 

The columns of data can be accessed from the Thermo object using indexing or the accessor syntax,

temp = thermo['Temp']
temp = thermo.Step

The latter can only be used if the columns are valid Python identifiers. To aid with this, the create_thermos function tidies up the column names.

Alternatively, the Thermo object has a .data attribute which contains the underlying Pandas DataFrame. Then any normal Pandas operations can be used on it.

By default only the last 'run' from the log file is loaded with th.create_thermos. Using the keyword arguments last=False and join=False will create a list of Thermo objects, one for each run. Using last=False and join=True will create a single Thermo object, with all the runs in a single file. last=True and join=True are the defaults.

Reading 'Chunk' Files

Chunks are spatial data outputted from the LAMMPS fix chunk/ave. This file contains multiple tables, each of which is a sub-average over the course of the simulation.

If the ave running argument is provided to fix chunk/ave, then only the last frame is needed. The th.create_chunk function reads just the last frame of this file.

  # Read data from chunk ave or RDF
  # Reads only the last table
  chunk = th.create_chunk('chunk.prof')
  plt.plot(chunk.Coord1,chunk.temp)
  plt.show()

If you need all the frames, then the th.create_multi_chunks function can create a MultiChunk object to load them all.

  # Read all the chunks from a chunk/ave file
  chunks = th.create_multi_chunks('chunk.prof')
  # Average the different frames of `MultiChunk` into a single `Chunk`
  chunk = chunks.flatten_chunk()

As with the Thermo class, Chunk and MultiChunk instances have a .data attribute, and can be indexed or use the getter syntax.

The Chunk format is also outputted by the LAMMPS command fix ave/time in vector mode.

The constructor th.create_chunk can also optionally read data outputted by the numpy function savetxt using the style='np' kwarg,

  chunk = th.create_chunk('data.txt',style='np')