bitWalk's: PySide6 と Matplotlib

2021-11-17

PySide6 と Matplotlib

Python / NumPy のためのグラフ描画ライブラリである Matplotlib 3.5.0 がリリースされ、PySide6 (Qt for Python) で利用できるようになりました。待ちに待ったリリースだったので、早速、過去のサンプル^[2]を手直しして実行できるようにしました。😁

本記事では、下記の OS 環境を使用しています。

Fedora 35 Workstation

x86_64

qt_matplotlib.py の実行例

PySide6 6.2.1
matplotlib 3.5.0

サンプルコードを下記に示しました。

qt_matplotlib.py

	#!/usr/bin/env python
	# coding: utf-8
	import PySide6
	from PySide6.QtWidgets import (
	QApplication,
	QVBoxLayout,
	QWidget
	)
	import sys
	import pandas as pd
	# Reference
	# https://qiita.com/hiroyuki_kageyama/items/cb87a0bee98c0262a35e
	import matplotlib

	matplotlib.use('Agg')
	import matplotlib.pyplot as plt
	from matplotlib.backends.backend_qtagg import FigureCanvasQTAgg as FigureCanvas
	from matplotlib.backends.backend_qtagg import NavigationToolbar2QT as NavigationToolbar


	class Example(QWidget):
	def __init__(self):
	super().__init__()
	self.setWindowTitle('SPC Chart Example')
	self.init_ui()
	print('PySide6', PySide6.__version__)
	print('matplotlib', matplotlib.__version__)

	def init_ui(self):
	# Example dataframe
	df = pd.DataFrame({
	'Sample': list(range(1, 11)),
	'Y': [9.030, 8.810, 9.402, 8.664, 8.773, 8.774, 8.416, 9.101, 8.687, 8.767]
	})
	# SPC metrics
	spec_usl = 9.97
	spec_target = 8.70
	spec_lsl = 7.43
	value_mean = df.describe().at['mean', 'Y']
	# _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_
	# SPC chart
	fig = plt.figure(dpi=100)
	ax = fig.add_subplot(111, title=self.windowTitle())
	plt.subplots_adjust(bottom=0.2, left=0.2, right=0.8, top=0.9)
	ax.grid(True)
	# Horizontal Lines for SPC metrics
	ax.axhline(y=spec_usl, linewidth=1, color='red', label='USL')
	ax.axhline(y=spec_target, linewidth=1, color='blue', label='Target')
	ax.axhline(y=spec_lsl, linewidth=1, color='red', label='LSL')
	ax.axhline(y=value_mean, linewidth=1, color='green', label='Avg')
	# Trend
	ax.plot(df['Sample'], df['Y'], color="gray", marker="o", markersize=10)
	ax.yaxis.label.set_color('gray')
	ax.tick_params(axis='y', colors='gray')
	# add extra ticks
	extraticks = [spec_lsl, spec_target, spec_usl]
	ax.set_yticks(list(ax.get_yticks()) + extraticks)
	fig.canvas.draw()
	# Labels
	labels = [item.get_text() for item in ax.get_yticklabels()]
	n = len(labels)
	labels[n - 3] = 'LSL = ' + str(spec_lsl)
	labels[n - 2] = 'Target = ' + str(spec_target)
	labels[n - 1] = 'USL = ' + str(spec_usl)
	ax.set_yticklabels(labels)
	# Color
	yticklabels = ax.get_yticklabels()
	n = len(yticklabels)
	yticklabels[n - 3].set_color('red')
	yticklabels[n - 2].set_color('blue')
	yticklabels[n - 1].set_color('red')
	# add second y axis wish same range as first y axis
	ax2 = ax.twinx()
	ax2.set_ylim(ax.get_ylim())
	ax2.tick_params(axis='y', colors='gray')
	# add extra ticks
	extraticks2 = [value_mean]
	ax2.set_yticks(list(ax2.get_yticks()) + extraticks2)
	# Label for second y axis
	labels2 = [item.get_text() for item in ax2.get_yticklabels()]
	n = len(labels2)
	labels2[n - 1] = 'Avg = ' + str(value_mean)
	ax2.set_yticklabels(labels2)
	# Color for second y axis
	yticklabels2 = ax2.get_yticklabels()
	n = len(yticklabels2)
	yticklabels2[n - 1].set_color('green')
	# _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_
	# Canvas & Toolbar
	canvas = FigureCanvas(fig)
	toolbar = NavigationToolbar(canvas, self)
	unwanted_buttons = ['Back', 'Forward']
	for x in toolbar.actions():
	if x.text() in unwanted_buttons:
	toolbar.removeAction(x)
	# _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_
	# Layout
	layout = QVBoxLayout(self)
	layout.setContentsMargins(0, 0, 0, 0)
	layout.setSpacing(0)
	layout.addWidget(toolbar)
	layout.addWidget(canvas)


	def main():
	app = QApplication(sys.argv)
	ex = Example()
	ex.show()
	sys.exit(app.exec())


	if __name__ == '__main__':
	main()