yu9824 / ternary_diagram Goto Github PK

View Code? Open in Web Editor NEW

14.0 1.0 1.0 17.82 MB

This package makes it easier for you to draw beautiful ternary diagram without pymatgen.

Home Page: https://yu9824.github.io/ternary_diagram/

License: MIT License

Python 95.69% Shell 0.53% Makefile 1.44% Batchfile 1.81% HTML 0.52%

ternary-diagrams chemistry materials-informatics matplotlib python visualization

ternary_diagram's Introduction

Ternary Diagram

This package makes it easier for you to draw beautiful ternary diagram without pymatgen.

Meaningly, only need numpy and matplotlib.

What you will be able to do with this package

Create beautiful contour maps easily
Creating a scatter plot
Draw tie lines
Automatically format chemical composition using subscripts
Most of the matplotlib options are available

How to install

PyPI

pip install ternary-diagram

PyPI project is here.

Anaconda

conda install -c conda-forge ternary-diagram

Anaconda (conda-forge) package site is here.

Usage

See Examples and the documentation.

Examples

An easy example is here.

import matplotlib.pyplot as plt
from ternary_diagram import TernaryDiagram

# You can set `ax` to select which axes to draw. If not, the current axes will be used.
td = TernaryDiagram(["Li2O", "La2O3", "TiO2"])

# scatter
td.scatter(vector=[[1, 1, 1], [1, 2, 3]], z=[0, 1])
# You can set some options in `plt.scatter` like `marker`, `c` etc.
td.scatter(vector=[[2, 1, 3], [3, 2, 1]], marker="s", c="#022c5e", s=30)

# line plot
# You can set some options in `plt.plot` like `lw`, `c`, and so on.
td.plot([[1, 1, 1], [1, 2, 3]], color="black")

# save figure
td.fig.savefig("figure.png", dpi=144)

It can be written like this.

# The background color is sometimes transparent in jupyter notebooks, so set facecolor 'white'.
fig, ax = plt.subplots(facecolor="w")

# You can set `ax` to select which axes to draw. If not, the current axes will be used.
td = TernaryDiagram(["Li2O", "La2O3", "TiO2"], ax=ax)

# scatter
td.scatter(vector=[[1, 1, 1], [1, 2, 3]], z=[0, 1])
# You can set some options in `plt.scatter` like `marker`, `c` etc.
td.scatter(vector=[[2, 1, 3], [3, 2, 1]], marker="s", c="#022c5e", s=30)

# line plot
# You can set some options in `plt.plot` like `lw`, `c`, and so on.
td.plot([[1, 1, 1], [1, 2, 3]], color="black")

# save figure
fig.savefig("figure.png", dpi=144)

It means that you can draw multiple figures in one figure object.

Histories

v3.1.0

Delete .utils._BasePlotter.get_x_y (we should access directly).
Resolve #10 (Create fill argument in TernaryDiagram.contour).
Resolve #11 (Create auto_latex_notation argument in TernaryDiagram).
Update documentation.
Update docstrings and typings.

LICENSE

See LICENSE.

MIT Licence

ternary_diagram's People

Contributors

Stargazers

Watchers

Forkers

francescapelusi

ternary_diagram's Issues

Do we need to manage the matplotlib.pyplot.figure object in TernaryDiagram object?

In seaborn, you have it out.
Maybe we should follow this approach.

Contour and Contourf

Shouldn't there be a distinction between contour and contourf?

Colorbar is not customizable.

Improve the ability to handle Colorber objects.

The error occurs when there is only one uniform value.

I believe that this problem stems from the fact, the upper and lower limits are the same value rather than the homogeneousness of data.

import numpy as np
from ternary_diagram import TernaryDiagram

td = TernaryDiagram(['Li2O', 'TiO2', 'La2O3'])

np.random.seed(334)
n_samples = 100
ratio = np.random.rand(n_samples, 3)
z = np.ones(n_samples)

td.contour(ratio, z=z)

Error message is here.

ValueError: Contour levels must be increasing
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-7-1baff06f772a> in <module>
     13     )
     14 
---> 15     module.predict(
     16         materials=['Li2O', 'SiO2', 'MoO3'],
     17         compositions=[],

/conductivity_prediction/main.py in predict(self, materials, compositions, path_dir)
    584 
    585             td = TernaryDiagram(materials = df_ratio.columns)
--> 586             td.contour(vector = df_ratio, z = y_pred)
    587             td.fig.savefig(os.path.join(path_output_img, '{0}.{1}'.format('-'.join(materials), 'png')), dpi = 300, transparent = False)
    588             plt.close()

/opt/conda/lib/python3.8/site-packages/ternary_diagram/ternary_diagram.py in contour(self, vector, **options)
    285             percentage of each compound mixed in 2D list / pandas.DataFrame / numpy.ndarray, where shape = [n, 3] (n is the number of samples to be plotted as integer)
    286         '''
--> 287         self._contour_(self, vector, **options) # selfオブジェクトを渡してる．
    288 
    289     def plot(self, r1, r2, **options):   # 連結線を引く (scatterオブジェクトの使用が必須な状況)

/opt/conda/lib/python3.8/site-packages/ternary_diagram/ternary_diagram.py in __init__(self, outer, vector, **options)
    238 
    239             T = tri.Triangulation(self.x, self.y)
--> 240             self.triplot = outer.ax.tricontourf(self.x, self.y, T.triangles, self.z, np.linspace(self.minimum if self.minimum is not None else np.min(self.z), self.maximum if self.maximum is not None else np.max(self.z), 101), **self.options)
    241             self.colorbar()
    242             self.tight_layout()

/opt/conda/lib/python3.8/site-packages/matplotlib/tri/tricontour.py in tricontourf(ax, *args, **kwargs)
    319     """
    320     kwargs['filled'] = True
--> 321     return TriContourSet(ax, *args, **kwargs)

/opt/conda/lib/python3.8/site-packages/matplotlib/tri/tricontour.py in __init__(self, ax, *args, **kwargs)
     38         are described in the docstring of `~.Axes.tricontour`.
     39         """
---> 40         ContourSet.__init__(self, ax, *args, **kwargs)
     41 
     42     def _process_args(self, *args, **kwargs):

/opt/conda/lib/python3.8/site-packages/matplotlib/contour.py in __init__(self, ax, levels, filled, linewidths, linestyles, hatches, alpha, origin, extent, cmap, colors, norm, vmin, vmax, extend, antialiased, nchunk, locator, transform, *args, **kwargs)
    814         self._transform = transform
    815 
--> 816         kwargs = self._process_args(*args, **kwargs)
    817         self._process_levels()
    818 

/opt/conda/lib/python3.8/site-packages/matplotlib/tri/tricontour.py in _process_args(self, *args, **kwargs)
     50         else:
     51             from matplotlib import _tri
---> 52             tri, z = self._contour_args(args, kwargs)
     53             C = _tri.TriContourGenerator(tri.get_cpp_triangulation(), z)
     54             self._mins = [tri.x.min(), tri.y.min()]

/opt/conda/lib/python3.8/site-packages/matplotlib/tri/tricontour.py in _contour_args(self, args, kwargs)
    105         if self.logscale and self.zmin <= 0:
    106             raise ValueError('Cannot %s log of negative values.' % fn)
--> 107         self._process_contour_level_args(args[1:])
    108         return (tri, z)
    109 

/opt/conda/lib/python3.8/site-packages/matplotlib/contour.py in _process_contour_level_args(self, args)
   1170 
   1171         if len(self.levels) > 1 and np.min(np.diff(self.levels)) <= 0.0:
-> 1172             raise ValueError("Contour levels must be increasing")
   1173 
   1174     def _process_levels(self):

ValueError: Contour levels must be increasing

Add options related to composition or not?

At present, the program automatically adds subscripts to the text entered as materials and various other text.

However, for extensibility, we think it should be possible to select the subscript by using flag.

For example,

td = TernaryDiagram(materials=['Li2O', 'La2O3', 'TiO2'], flag_composition = False)

etc.

For td.annotate in particular, we think it should be recognized as normal text by default. In particular, for td.annotate, we think it should be recognized as normal text by default, because it is easy to create subscripts using ternary_diagram.utils.get_label function.

Modify error about contour function.

from ternary_diagram import TernaryDiagram
import numpy as np

ratio = np.random.rand(100, 3)

td = TernaryDiagram(materials = ['Li2O', 'TiO2', 'La2O3'])
td.contour(ratio)

Error statement here.

AttributeError: '_contour_' object has no attribute 'minimum'
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-3-a6f20240bea1> in <module>
      1 td = TernaryDiagram(materials = ['Li2O', 'TiO2', 'La2O3'])
----> 2 td.contour(ratio)

~/ternary_diagram/ternary_diagram/ternary_diagram.py in contour(self, vector, **options)
    285             percentage of each compound mixed in 2D list / pandas.DataFrame / numpy.ndarray, where shape = [n, 3] (n is the number of samples to be plotted as integer)
    286         '''
--> 287         self._contour_(self, vector, **options) # selfオブジェクトを渡してる．
    288 
    289     def plot(self, r1, r2, **options):   # 連結線を引く (scatterオブジェクトの使用が必須な状況)

~/ternary_diagram/ternary_diagram/ternary_diagram.py in __init__(self, outer, vector, **options)
    238 
    239             T = tri.Triangulation(self.x, self.y)
--> 240             self.triplot = outer.ax.tricontourf(self.x, self.y, T.triangles, self.z, np.linspace(self.minimum if self.minimum is not None else np.min(self.z), self.maximum if self.maximum is not None else np.max(self.z), 101), **self.options)
    241             self.colorbar()
    242             self.tight_layout()

AttributeError: '_contour_' object has no attribute 'minimum'

You should encourage the programmer to improve by issuing an error message "You must enter 'z'." if he does not enter 'z' when the contour function is called.