Pandas 如何使用matplotlib和pyplot正确实现QThread
我知道还有一两个问题是相关的,但不完全是我需要的。我正在构建这个gui,它通过单击一个按钮来激活一个模块。这个通过按下按钮激活的python模块从多个pandas数据帧生成热图并保存这些图像,然后使用pandas ExcelWriter将这些图像保存到xlsx中 我曾尝试实现QThread,正如其他stackoverflow示例试图解释类似的问题一样,但我继续遇到以下错误:Pandas 如何使用matplotlib和pyplot正确实现QThread,pandas,matplotlib,pyqt4,seaborn,qthread,Pandas,Matplotlib,Pyqt4,Seaborn,Qthread,我知道还有一两个问题是相关的,但不完全是我需要的。我正在构建这个gui,它通过单击一个按钮来激活一个模块。这个通过按下按钮激活的python模块从多个pandas数据帧生成热图并保存这些图像,然后使用pandas ExcelWriter将这些图像保存到xlsx中 我曾尝试实现QThread,正如其他stackoverflow示例试图解释类似的问题一样,但我继续遇到以下错误:“在GUI线程之外使用pixmaps是不安全的”。我知道从技术上讲,我不是在主gui线程内创建热图,但我认为使用QThrea
“在GUI线程之外使用pixmaps是不安全的”
。我知道从技术上讲,我不是在主gui线程内创建热图,但我认为使用QThread,我仍然在“a”gui线程内。热图所基于的这些数据帧有时会很大,我在某种程度上掌握了在创建热图时向主gui线程发送信号的概念,并且在主gui类中具有热图功能……但我担心这会在以后传递如此多的数据时带来麻烦……这更像流水线比线程更重要。我只希望这个工作线程创建这些图像并保存它们,然后将这些保存的文件保存到xlsx中,而不中断主gui
(注意:这是一个简化版本,在实际程序中,将几乎同时创建多个线程,每个线程内将创建多个热图)
---main.py---
---excel_dummy.py---
---MAIN_GUI.py---
即使您明确使用
Agg
后端生成您的体形,看起来Seaborn
仍在使用系统上的默认后端,这很可能是一个交互式后端Qt4Agg
。我们希望Seaborn使用非交互式后端
,以避免任何错误(有关后端的更多详细信息,请参阅)。为此,请在导入中告诉Matplotlib
使用Agg
后端,并在Matplotlib之后导入Seaborn
您还需要将图形保存为png,因为Agg
后端不支持jpg。除非您有使用jpg的特定原因,否则png通常是图形的更好格式
最后,您可以使用内存缓冲区,而不是在将图像保存到Excel工作簿之前将其保存到临时文件中。我还没有测试过它,但是如果您使用的是大文件,它可能会更快
下面是我编写的一个MWE,其中包括上述要点,并且在Python3.4中没有给出我的系统上的任何错误:
import pandas as pd
import time
from pandas import ExcelWriter
import numpy as np
from PyQt4 import QtCore, QtGui
import matplotlib as mpl
mpl.use('Agg')
from matplotlib.backends.backend_agg import FigureCanvas
import seaborn.matrix as sm
try: # Python 2 (not tested)
from cStringIO import StringIO as BytesIO
except ImportError: # Python 3
from io import BytesIO
class MAIN_GUI(QtGui.QWidget):
def __init__(self):
super(MAIN_GUI, self).__init__()
self.worker = Excelify()
btn = QtGui.QPushButton('Run')
disp = QtGui.QLabel()
self.setLayout(QtGui.QGridLayout())
self.layout().addWidget(btn, 0, 0)
self.layout().addWidget(disp, 2, 0)
self.layout().setRowStretch(1, 100)
btn.clicked.connect(self.worker.start)
self.worker.figSaved.connect(disp.setText)
class Excelify(QtCore.QThread):
figSaved = QtCore.pyqtSignal(str)
def run(self):
self.figSaved.emit('Saving figure to Workbook.')
t1 = time.clock()
image_data = self.heatmap()
with ExcelWriter('final.xlsx', engine='xlsxwriter') as writer:
wb = writer.book
ws = wb.add_worksheet()
ws.insert_image('C3', 'heat.png', {'image_data': image_data})
writer.save()
t2 = time.clock()
self.figSaved.emit('Done in %f sec.' % (t2-t1))
def heatmap(self):
df = pd.DataFrame(np.array([[1, 22222, 33333], [2, 44444, 55555],
[3, 44444, 22222], [4, 55555, 33333]]),
columns=['hour', 'in', 'out'])
dfu = pd.DataFrame(df.groupby([df.out, df.hour]).size())
dfu.reset_index(inplace=True)
dfu.rename(columns={'0': 'Count'})
dfu.columns = ['in', 'hour', 'Count']
fig = mpl.figure.Figure()
fig.set_canvas(FigureCanvas(fig))
ax = fig.add_subplot(111)
df_heatmap = dfu.pivot('in', 'hour', 'Count').fillna(0)
sm.heatmap(df_heatmap, ax=ax, square=True, annot=False, mask=0)
buf= BytesIO()
fig.savefig(buf, format='png')
return(buf)
if __name__ == '__main__':
import sys
app = QtGui.QApplication(sys.argv)
w = MAIN_GUI()
w.show()
w.setFixedSize(200, 100)
sys.exit(app.exec_())
import os, pandas as pd
from pandas import ExcelWriter
import numpy as np
import seaborn.matrix as sm
from PyQt4 import QtCore
from PyQt4.QtCore import QThread
from matplotlib.backends.backend_agg import FigureCanvas
from matplotlib.figure import Figure
import time
class excelify(QThread):
def __init__(self):
QThread.__init__(self)
def run(self):
path = 'home/desktop/produced_files'
with ExcelWriter(path + '/final.xlsx', engine='xlsxwriter') as writer:
workbook = writer.book
worksheet = workbook.add_worksheet()
heatit = self.heatmap()
worksheet.insert_image('C3',path + '/' + 'heat.jpg')
worksheet.write(2, 2, 'just write something')
writer.save()
print('file size: %s "%s"' % (os.stat(path).st_size, path))
time.slee(0.3)
self.emit(QtCore.SIGNAL('donethread(QString)'),'')
def heatmap(self):
df = pd.DataFrame(np.array([[1,22222,33333],[2,44444,55555],[3,44444,22222],[4,55555,33333]]),columns=['hour','in','out'])
dfu = pd.DataFrame(df.groupby([df.in,df.hour]).size())
dfu.reset_index(inplace=True)
dfu.rename(columns={'0':'Count'})
dfu.columns=['in','hour','Count']
dfu_2 = dfu.copy()
mask=0
fig = Figure()
ax = fig.add_subplot(1,1,1)
canvas = FigureCanvas(fig)
df_heatmap = dfu_2.pivot('in','hour','Count').fillna(0)
sm.heatmap(df_heatmap,ax=ax,square=True,annot=False,mask=mask)
fig.savefig(path + '/' + heat.jpg')
from PyQt4 import QtCore,QtGui
try:
_fromUtf8 = QtCore.QString.fromUtf8
except AttributeError:
def _fromUtf8(s):
return s
try:
_encoding = QtGui.QApplication.unicodeUTF8
def _translate(context, text, disambig):
return QtGui.QApplication.translate(context, text, disambig, _encoding)
except AttributeError:
def _translate(context, text, disambig):
return QtGui.QApplication.translate(context, text, disambig)
class Ui_MainWindow(object):
def setupUi(self, MainWindow):
MainWindow.setObjectName(_fromUtf8("MainWindow"))
MainWindow.resize(320,201)
self.centralwidget = QtGui.QWidget(MainWindow)
self.centralwidget.setObjectName(_fromUtf8("centralwidget"))
self.updateALL_Button = QtGui.QPushButton(self.centralwidget)
self.updateALL_Button.setGeometry(QtCore.QRect(40,110,161,27))
self.updateALL_Button.setFocusPolicy(QtCore.Qt.NoFocus)
self.updateALL_Button.setObjectName(_fromUtf8("Options_updateALL_Button"))
MainWindow.setCentralWidget(self.centralwidget)
self.menubar = QtGui.QMenuBar(MainWindow)
self.menubar.setGeometry(QtCore.QRect(0, 0, 320, 24))
self.menubar.setObjectName(_fromUtf8("menubar"))
MainWindow.setMenuBar(self.menubar)
self.statusbar = QtGui.QStatusBar(MainWindow)
self.statusbar.setObjectName(_fromUtf8("statusbar"))
MainWindow.setStatusBar(self.statusbar)
self.retranslateUi(MainWindow)
QtCore.QMetaObject.connectSlotsByName(MainWindow)
def retranslateUi(self,MainWindow):
MainWindow.setWindowTitle(_translate("MainWindow", "MainWindow", None))
self.updateALL_Button.setText(_translate("MainWindow", "updateALL", None))
import pandas as pd
import time
from pandas import ExcelWriter
import numpy as np
from PyQt4 import QtCore, QtGui
import matplotlib as mpl
mpl.use('Agg')
from matplotlib.backends.backend_agg import FigureCanvas
import seaborn.matrix as sm
try: # Python 2 (not tested)
from cStringIO import StringIO as BytesIO
except ImportError: # Python 3
from io import BytesIO
class MAIN_GUI(QtGui.QWidget):
def __init__(self):
super(MAIN_GUI, self).__init__()
self.worker = Excelify()
btn = QtGui.QPushButton('Run')
disp = QtGui.QLabel()
self.setLayout(QtGui.QGridLayout())
self.layout().addWidget(btn, 0, 0)
self.layout().addWidget(disp, 2, 0)
self.layout().setRowStretch(1, 100)
btn.clicked.connect(self.worker.start)
self.worker.figSaved.connect(disp.setText)
class Excelify(QtCore.QThread):
figSaved = QtCore.pyqtSignal(str)
def run(self):
self.figSaved.emit('Saving figure to Workbook.')
t1 = time.clock()
image_data = self.heatmap()
with ExcelWriter('final.xlsx', engine='xlsxwriter') as writer:
wb = writer.book
ws = wb.add_worksheet()
ws.insert_image('C3', 'heat.png', {'image_data': image_data})
writer.save()
t2 = time.clock()
self.figSaved.emit('Done in %f sec.' % (t2-t1))
def heatmap(self):
df = pd.DataFrame(np.array([[1, 22222, 33333], [2, 44444, 55555],
[3, 44444, 22222], [4, 55555, 33333]]),
columns=['hour', 'in', 'out'])
dfu = pd.DataFrame(df.groupby([df.out, df.hour]).size())
dfu.reset_index(inplace=True)
dfu.rename(columns={'0': 'Count'})
dfu.columns = ['in', 'hour', 'Count']
fig = mpl.figure.Figure()
fig.set_canvas(FigureCanvas(fig))
ax = fig.add_subplot(111)
df_heatmap = dfu.pivot('in', 'hour', 'Count').fillna(0)
sm.heatmap(df_heatmap, ax=ax, square=True, annot=False, mask=0)
buf= BytesIO()
fig.savefig(buf, format='png')
return(buf)
if __name__ == '__main__':
import sys
app = QtGui.QApplication(sys.argv)
w = MAIN_GUI()
w.show()
w.setFixedSize(200, 100)
sys.exit(app.exec_())