pydiagram 0.6

1. Install

$ pip install pydiagram

Required Packages

• numpy

• scipy

• matplotlib

• mpltex

• attrdict

• PyYAML

Important note: PyYAML cannot be installed by pip or easy_install. Please install it by downloading its source code and use python setup.py install to install it, see details at PyYAML Documentation.

2. Use PyDiagram as An Executable

To use the executable, pydiagram, you need to provide a project configuration file. The configuration file is in YAML format. Three sample configuration files, config_polyorder.yml, config_polyfts.yml, and config_pydiagram.yml, are shipped with PyDiagram package. You can copy one of them to your project root directory and make any necessary modifications to it.

Typical usage.

pydiagram [-v] [-h] [-p] [-q] [-s] [-m]

• pydiagram -p

This command processes the simulation data in the current directory. It will save diagram, info_map, and boundary objects as Python pickle files. The parser can be specified in the project configuration file using the solver option.

• pydiagram -q

According to the plot mode specified in the project configuration file, this command will plot requested figures and save them as files.

• pydiagram -s

This command will run pydiagram as a server.

• pydiagram -m

This command will run pydiagram in manual mode, which will submit additional simulation jobs specified in the predictor.manual in the project configuration file.

3. Use PyDiagram as A Library

You can customize your own processing, plotting and predicting scripts by utilizing PyDiagram functions. As described in the last section, your script will also depend on a project configuration file.

Typical usage.

import pydiagram
import matplotlib.pyplot as plt

# Generate diagram by specifying the folder where data locate.
# This will require the output data files from supported solvers.
# Here assume the project configuration file in the current directory.
diagram = pydiagram.get_diagram()
# Plot the raw diagram
fig, ax = plt.subplots(1)
pydiagram.plot_diagram(ax, diagram)

# Or you can generate diagram from .dgm file
# First load .dgm file to get info_map
xaxis, yaxis, info_map = pydiagram.load_dgm('phase_data.dgm')
diagram = pydiagram.get_diagram_from_info_map(info_map)

# You can find phase boundary from digaram
boundary = pydiagram.get_boundary(diagram, info_map)
# And plot it as scatter points
pydiagram.plot_boundary_point(ax, boundary)
# Or plot it as line
pydiagram.plot_boundary_line(ax, boundary)

# PyDiagram also provides a plot function for plotting diagram points
# and phase boundary together.
pydiagram.plot_phase_diagram(diagram, boundary, settings, xaxis, yaxis)

# To analyze the simulation results, PyDiagram also provides
# a plotting function which will plot F vs. x or F vs. y,
# a vs. x or a vs. y, and accuracy vs. x or accuracy vs. y.
pydiagram(info_map, xaxis, yaxis, path, phases, val)

All above mentioned functions have more available options to fine tune the processing and plotting. For more details, please consult the source code or directly contact me via E-mail.

4. dgm File Format

For non Polyorder or PolyFTS users, to use PyDiagram you have to generate your own dgm file from your simulation results by following the dgm file format.

The format of the dgm file.

First line:

[x-axis name] [y-axis name]

Other lines:

[x] [y] [phase] [F] [a] [accuracy] [separation state]

• x-axis name: the variable name of the x axis in the phase diagram.

• y-axis name: the variable name of the y axis in the phase diagram.

• x and y: the coordinates in the phase diagram.

• phase: the name of the phase structure.

• F: the Gibbs free energy of the corresponding phase structure.

• a: the stretch-free cell size.

• accuracy: the residual error for the corresponding simulation.

• separation state: Does the simulation yield phase separated structures? 0 for no and 1 for yes.

The first five columns are mandatory while the last two are optional. The optional values are assumed by their order. For example, if only 6 columns are provided, then the 6th column is interpreted as accuracy.

Example:

f   xN
0.1 12.5    HEX 3.953993    2.9543  8.7e-7  1
0.1 12.6    LAM 3.970134    2.5877  1.7e-8  1