blocks|key|316035|text|一种简单的方法可能是只编写包含gnuplot命令的第三个文件，然后告诉Python使用该文件执行gnuplot。假设你写了|type|unstyled|depth|inlineStyleRanges|entityRanges|data|316036|"plot+'%25s'+with+lines,+'%25s'+with+points;"+%25+(eout,nout)|code-block|syntax|javascript|316037|添加到名为tmp.gp的文件中。然后你就可以使用|316038|from+os+import+system,+remove
system('gnuplot+-persist+tmp.gp')
remove('tmp.gp')|316039|entityMap^0|0|0|0|0^^$0|@$1|2|3|4|5|6|7|M|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|N|8|@]|9|@]|A|$E|F]]|$1|G|3|H|5|6|7|O|8|@]|9|@]|A|$]]|$1|I|3|J|5|D|7|P|8|@]|9|@]|A|$E|F]]|$1|K|3|-4|5|6|7|Q|8|@]|9|@]|A|$]]]|L|$]]

A simple approach might be to just write a third file containing your gnuplot commands and then tell Python to execute gnuplot with that file. Say you write
<pre><code>&quot;plot '%s' with lines, '%s' with points;&quot; % (eout,nout)
</code></pre>
to a file called tmp.gp. Then you can use
<pre><code>from os import system, remove
system('gnuplot -persist tmp.gp')
remove('tmp.gp')
</code></pre>

blocks|key|3065265|text|subprocess模块允许您调用其他程序：|type|unstyled|depth|inlineStyleRanges|offset|length|style|CODE|entityRanges|data|3065266|import+subprocess
plot+=+subprocess.Popen(['gnuplot'],+stdin=subprocess.PIPE)
plot.communicate("plot+'%25s'+with+lines,+'%25s'+with+points;"+%25+(eout,nout))|code-block|syntax|javascript|3065267|entityMap|0|LINK|mutability|MUTABLE|url|http://docs.python.org/library/subprocess.html^0|0|A|0|A|0|0|0^^$0|@$1|2|3|4|5|6|7|S|8|@$9|T|A|U|B|C]]|D|@$9|V|A|W|1|X]]|E|$]]|$1|F|3|G|5|H|7|Y|8|@]|D|@]|E|$I|J]]|$1|K|3|-4|5|6|7|Z|8|@]|D|@]|E|$]]]|L|$M|$5|N|O|P|E|$Q|R]]]]

The <a href="http://docs.python.org/library/subprocess.html" rel="noreferrer"><code>subprocess</code></a> module lets you call other programs:

<pre><code>import subprocess
plot = subprocess.Popen(['gnuplot'], stdin=subprocess.PIPE)
plot.communicate("plot '%s' with lines, '%s' with points;" % (eout,nout))
</code></pre>

blocks|key|3278338|text|子过程在Doug+Hellemann的Python+Module+of+the+Week中解释得非常清楚|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|3278339|这样做效果很好：|3278340|import+subprocess
proc+=+subprocess.Popen(['gnuplot','-p'],+
++++++++++++++++++++++++shell=True,
++++++++++++++++++++++++stdin=subprocess.PIPE,
++++++++++++++++++++++++)
proc.stdin.write('set+xrange+[0:10];+set+yrange+[-2:2]\n')
proc.stdin.write('plot+sin(x)\n')
proc.stdin.write('quit\n')+#close+the+gnuplot+window|code-block|syntax|javascript|3278341|也可以使用'communicate‘，但除非使用gnuplot+pause命令，否则绘图窗口会立即关闭|3278342|proc.communicate("""
set+xrange+[0:10];+set+yrange+[-2:2]
plot+sin(x)
pause+4
""")|3278343|entityMap|0|LINK|mutability|MUTABLE|url|http://www.doughellmann.com/PyMOTW/subprocess/^0|J|P|0|0|0|0|0|0^^$0|@$1|2|3|4|5|6|7|W|8|@]|9|@$A|X|B|Y|1|Z]]|C|$]]|$1|D|3|E|5|6|7|10|8|@]|9|@]|C|$]]|$1|F|3|G|5|H|7|11|8|@]|9|@]|C|$I|J]]|$1|K|3|L|5|6|7|12|8|@]|9|@]|C|$]]|$1|M|3|N|5|H|7|13|8|@]|9|@]|C|$I|J]]|$1|O|3|-4|5|6|7|14|8|@]|9|@]|C|$]]]|P|$Q|$5|R|S|T|C|$U|V]]]]

Subprocess is explained very clearly on Doug Hellemann's 
<a href="http://www.doughellmann.com/PyMOTW/subprocess/" rel="noreferrer">Python Module of the Week</a>

This works well:

<pre><code>import subprocess
proc = subprocess.Popen(['gnuplot','-p'], 
 shell=True,
 stdin=subprocess.PIPE,
 )
proc.stdin.write('set xrange [0:10]; set yrange [-2:2]\n')
proc.stdin.write('plot sin(x)\n')
proc.stdin.write('quit\n') #close the gnuplot window
</code></pre>

One could also use 'communicate' but the plot window closes immediately unless a gnuplot pause command is used

<pre><code>proc.communicate("""
set xrange [0:10]; set yrange [-2:2]
plot sin(x)
pause 4
""")
</code></pre>

blocks|key|3065352|text|我试图做一些类似的事情，但另外我想从python中提供数据，并将图形文件作为变量输出(因此，数据和图形都不是实际的文件)。这是我想出来的：|type|unstyled|depth|inlineStyleRanges|entityRanges|data|3065353|#!+/usr/bin/env+python

import+subprocess
from+sys+import+stdout,+stderr
from+os+import+linesep+as+nl

def+gnuplot_ExecuteCommands(commands,+data):
++++args+=+["gnuplot",+"-e",+(";".join([str(c)+for+c+in+commands]))]
++++program+=+subprocess.Popen(\
++++++++args,+\
++++++++stdin=subprocess.PIPE,+\
++++++++stdout=subprocess.PIPE,+\
++++++++stderr=subprocess.PIPE,+\
++++++++)
++++for+line+in+data:
++++++++program.stdin.write(str(line)%2Bnl)
++++return+program

def+gnuplot_GifTest():
++++commands+=+[\
++++++++"set+datafile+separator+','",\
++++++++"set+terminal+gif",\
++++++++"set+output",\
++++++++"plot+'-'+using+1:2+with+linespoints,+''+using+1:2+with+linespoints",\
++++++++]
++++data+=+[\
++++++++"1,1",\
++++++++"2,2",\
++++++++"3,5",\
++++++++"4,2",\
++++++++"5,1",\
++++++++"e",\
++++++++"1,5",\
++++++++"2,4",\
++++++++"3,1",\
++++++++"4,4",\
++++++++"5,5",\
++++++++"e",\
++++++++]
++++return+(commands,+data)

if+__name__=="__main__":
++++(commands,+data)+=+gnuplot_GifTest()
++++plotProg+=+gnuplot_ExecuteCommands(commands,+data)
++++(out,+err)+=+(plotProg.stdout,+plotProg.stderr)
++++stdout.write(out.read())|code-block|syntax|javascript|3065354|该脚本将图形转储到标准输出，作为main中的最后一步。等效的命令行(其中图形通过管道传输到'out.gif')将是：|offset|length|style|BOLD|3065355|gnuplot+-e+"set+datafile+separator+',';+set+terminal+gif;+set+output;+plot+'-'+using+1:2+with+linespoints,+''+using+1:2+with+linespoints"+>+out.gif
1,1
2,2
3,5
4,2
5,1
e
1,5
2,4
3,1
4,4
5,5
e|3065356|entityMap^0|0|0|G|4|0|0^^$0|@$1|2|3|4|5|6|7|Q|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|R|8|@]|9|@]|A|$E|F]]|$1|G|3|H|5|6|7|S|8|@$I|T|J|U|K|L]]|9|@]|A|$]]|$1|M|3|N|5|D|7|V|8|@]|9|@]|A|$E|F]]|$1|O|3|-4|5|6|7|W|8|@]|9|@]|A|$]]]|P|$]]

I was trying to do something similar, but additionally I wanted to feed data from within python and output the graph file as a variable (so neither the data nor the graph are actual files). This is what I came up with:

<pre><code>#! /usr/bin/env python

import subprocess
from sys import stdout, stderr
from os import linesep as nl

def gnuplot_ExecuteCommands(commands, data):
 args = ["gnuplot", "-e", (";".join([str(c) for c in commands]))]
 program = subprocess.Popen(\
 args, \
 stdin=subprocess.PIPE, \
 stdout=subprocess.PIPE, \
 stderr=subprocess.PIPE, \
 )
 for line in data:
 program.stdin.write(str(line)+nl)
 return program

def gnuplot_GifTest():
 commands = [\
 "set datafile separator ','",\
 "set terminal gif",\
 "set output",\
 "plot '-' using 1:2 with linespoints, '' using 1:2 with linespoints",\
 ]
 data = [\
 "1,1",\
 "2,2",\
 "3,5",\
 "4,2",\
 "5,1",\
 "e",\
 "1,5",\
 "2,4",\
 "3,1",\
 "4,4",\
 "5,5",\
 "e",\
 ]
 return (commands, data)

if __name__=="__main__":
 (commands, data) = gnuplot_GifTest()
 plotProg = gnuplot_ExecuteCommands(commands, data)
 (out, err) = (plotProg.stdout, plotProg.stderr)
 stdout.write(out.read())
</code></pre>

That script dumps the graph to stdout as the last step in main. The equivalent command line (where the graph is piped to 'out.gif') would be:

<pre><code>gnuplot -e "set datafile separator ','; set terminal gif; set output; plot '-' using 1:2 with linespoints, '' using 1:2 with linespoints" &gt; out.gif
1,1
2,2
3,5
4,2
5,1
e
1,5
2,4
3,1
4,4
5,5
e
</code></pre>

blocks|key|321099|text|我接受了Ben的建议，因为我在计算芹菜作业的图表时，发现从stdout读取时会锁定。我将其重新设计为使用StringIO创建指定给标准输入和subprocess.communicate的文件，以便通过标准输出立即获得结果，不需要读取。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|321100|from+subprocess+import+Popen,+PIPE
from+StringIO+import+StringIO++++++++++++++++++++++++++++++++++++++++++++
from+os+import+linesep+as+nl

def+gnuplot(commands,+data):++++++++++++++++++++++++++++++++++++++++++++++++++++
++++"""+drive+gnuplot,+expects+lists,+returns+stdout+as+string+"""++++++++++++++

++++dfile+=+StringIO()++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
++++for+line+in+data:+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
++++++++dfile.write(str(line)+%2B+nl)+++++++++++++++++++++++++++++++++++++++++++++

++++args+=+["gnuplot",+"-e",+(";".join([str(c)+for+c+in+commands]))]++++++++++++
++++p+=+Popen(args,+stdin=PIPE,+stdout=PIPE,+stderr=PIPE)+++++++++++++++++++++++

++++dfile.seek(0)+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
++++return+p.communicate(dfile.read())[0]+++

def+gnuplot_GifTest():
++++commands+=+[\
++++++++"set+datafile+separator+','",\
++++++++"set+terminal+gif",\
++++++++"set+output",\
++++++++"plot+'-'+using+1:2+with+linespoints,+''+using+1:2+with+linespoints",\
++++++++]
++++data+=+[\
++++++++"1,1",\
++++++++"2,2",\
++++++++"3,5",\
++++++++"4,2",\
++++++++"5,1",\
++++++++"e",\
++++++++"1,5",\
++++++++"2,4",\
++++++++"3,1",\
++++++++"4,4",\
++++++++"5,5",\
++++++++"e",\
++++++++]
++++return+(commands,+data)

if+__name__=="__main__":
++++(commands,+data)+=+gnuplot_GifTest()
++++print+gnuplot(commands,+data)|code-block|syntax|javascript|321101|entityMap^0|0|0^^$0|@$1|2|3|4|5|6|7|I|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|J|8|@]|9|@]|A|$E|F]]|$1|G|3|-4|5|6|7|K|8|@]|9|@]|A|$]]]|H|$]]

I went with Ben's suggestion as I was computing charts from a celery job and found that it would lockup when reading from stdout. I redesigned it like so using StringIO to create the file destined for stdin and subprocess.communicate to get the result immediately via stdout, no read required.

<pre><code>
from subprocess import Popen, PIPE
from StringIO import StringIO 
from os import linesep as nl

def gnuplot(commands, data): 
 """ drive gnuplot, expects lists, returns stdout as string """ 

 dfile = StringIO() 
 for line in data: 
 dfile.write(str(line) + nl) 

 args = ["gnuplot", "-e", (";".join([str(c) for c in commands]))] 
 p = Popen(args, stdin=PIPE, stdout=PIPE, stderr=PIPE) 

 dfile.seek(0) 
 return p.communicate(dfile.read())[0] 

def gnuplot_GifTest():
 commands = [\
 "set datafile separator ','",\
 "set terminal gif",\
 "set output",\
 "plot '-' using 1:2 with linespoints, '' using 1:2 with linespoints",\
 ]
 data = [\
 "1,1",\
 "2,2",\
 "3,5",\
 "4,2",\
 "5,1",\
 "e",\
 "1,5",\
 "2,4",\
 "3,1",\
 "4,4",\
 "5,5",\
 "e",\
 ]
 return (commands, data)

if __name__=="__main__":
 (commands, data) = gnuplot_GifTest()
 print gnuplot(commands, data)
</code></pre>

blocks|key|318123|text|下面是一个为wgnuplot.exe提供接口的类：|type|unstyled|depth|inlineStyleRanges|entityRanges|data|318124|from+ctypes+import+*
import+time
import+sys
import+os

#
#+some+win32+constants
#
WM_CHAR+++++=+0X0102
WM_CLOSE++++=+16
SW_HIDE+++++=+0
STARTF_USESHOWWINDOW+=+1

WORD++++=+c_ushort
DWORD+++=+c_ulong
LPBYTE++=+POINTER(c_ubyte)
LPTSTR++=+POINTER(c_char)+
HANDLE++=+c_void_p

class+STARTUPINFO(Structure):
++++_fields_+=+[("cb",DWORD),
++++++++("lpReserved",LPTSTR),+
++++++++("lpDesktop",+LPTSTR),
++++++++("lpTitle",+LPTSTR),
++++++++("dwX",+DWORD),
++++++++("dwY",+DWORD),
++++++++("dwXSize",+DWORD),
++++++++("dwYSize",+DWORD),
++++++++("dwXCountChars",+DWORD),
++++++++("dwYCountChars",+DWORD),
++++++++("dwFillAttribute",+DWORD),
++++++++("dwFlags",+DWORD),
++++++++("wShowWindow",+WORD),
++++++++("cbReserved2",+WORD),
++++++++("lpReserved2",+LPBYTE),
++++++++("hStdInput",+HANDLE),
++++++++("hStdOutput",+HANDLE),
++++++++("hStdError",+HANDLE),]

class+PROCESS_INFORMATION(Structure):
++++_fields_+=+[("hProcess",+HANDLE),
++++++++("hThread",+HANDLE),
++++++++("dwProcessId",+DWORD),
++++++++("dwThreadId",+DWORD),]

#
#+Gnuplot
#
class+Gnuplot:
++++#
++++#+__init__
++++#
++++def+__init__(self,+path_to_exe):
++++++++#+open+gnuplot
++++++++self.launch(path_to_exe)
++++++++#+wait+till+it's+ready
++++++++if(windll.user32.WaitForInputIdle(self.hProcess,+1000)):
++++++++++++print+"Error:+Gnuplot+timeout!"
++++++++++++sys.exit(1)
++++++++#+get+window+handles
++++++++self.hwndParent+=+windll.user32.FindWindowA(None,+'gnuplot')
++++++++self.hwndText+=+windll.user32.FindWindowExA(self.hwndParent,+None,+'wgnuplot_text',+None)



++++#
++++#+__del__
++++#
++++def+__del__(self):
++++++++windll.kernel32.CloseHandle(self.hProcess);
++++++++windll.kernel32.CloseHandle(self.hThread);
++++++++windll.user32.PostMessageA(self.hwndParent,+WM_CLOSE,+0,+0)


++++#
++++#+launch
++++#
++++def+launch(self,+path_to_exe):
++++++++startupinfo+=+STARTUPINFO()
++++++++process_information+=+PROCESS_INFORMATION()

++++++++startupinfo.dwFlags+=+STARTF_USESHOWWINDOW
++++++++startupinfo.wShowWindow+=+SW_HIDE

++++++++if+windll.kernel32.CreateProcessA(path_to_exe,+None,+None,+None,+False,+0,+None,+None,+byref(startupinfo),+byref(process_information)):
++++++++++++self.hProcess+=+process_information.hProcess
++++++++++++self.hThread+=+process_information.hThread
++++++++else:
++++++++++++print+"Error:+Create+Process+-+Error+code:+",+windll.kernel32.GetLastError()
++++++++++++sys.exit(1)



++++#
++++#+execute
++++#
++++def+execute(self,+script,+file_path):
++++++++#+make+sure+file+doesn't+exist
++++++++try:+os.unlink(file_path)
++++++++except:+pass

++++++++#+send+script+to+gnuplot+window
++++++++for+c+in+script:+windll.user32.PostMessageA(self.hwndText,+WM_CHAR,+ord(c),+1L)

++++++++#+wait+till+gnuplot+generates+the+chart
++++++++while(+not+(os.path.exists(file_path)+and+(os.path.getsize(file_path)+>+0))):+time.sleep(0.01)|code-block|syntax|javascript|318125|entityMap^0|0|0^^$0|@$1|2|3|4|5|6|7|I|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|J|8|@]|9|@]|A|$E|F]]|$1|G|3|-4|5|6|7|K|8|@]|9|@]|A|$]]]|H|$]]

Here's a class that provides an interface to wgnuplot.exe:

<pre><code>from ctypes import *
import time
import sys
import os

#
# some win32 constants
#
WM_CHAR = 0X0102
WM_CLOSE = 16
SW_HIDE = 0
STARTF_USESHOWWINDOW = 1

WORD = c_ushort
DWORD = c_ulong
LPBYTE = POINTER(c_ubyte)
LPTSTR = POINTER(c_char) 
HANDLE = c_void_p

class STARTUPINFO(Structure):
 _fields_ = [("cb",DWORD),
 ("lpReserved",LPTSTR), 
 ("lpDesktop", LPTSTR),
 ("lpTitle", LPTSTR),
 ("dwX", DWORD),
 ("dwY", DWORD),
 ("dwXSize", DWORD),
 ("dwYSize", DWORD),
 ("dwXCountChars", DWORD),
 ("dwYCountChars", DWORD),
 ("dwFillAttribute", DWORD),
 ("dwFlags", DWORD),
 ("wShowWindow", WORD),
 ("cbReserved2", WORD),
 ("lpReserved2", LPBYTE),
 ("hStdInput", HANDLE),
 ("hStdOutput", HANDLE),
 ("hStdError", HANDLE),]

class PROCESS_INFORMATION(Structure):
 _fields_ = [("hProcess", HANDLE),
 ("hThread", HANDLE),
 ("dwProcessId", DWORD),
 ("dwThreadId", DWORD),]

#
# Gnuplot
#
class Gnuplot:
 #
 # __init__
 #
 def __init__(self, path_to_exe):
 # open gnuplot
 self.launch(path_to_exe)
 # wait till it's ready
 if(windll.user32.WaitForInputIdle(self.hProcess, 1000)):
 print "Error: Gnuplot timeout!"
 sys.exit(1)
 # get window handles
 self.hwndParent = windll.user32.FindWindowA(None, 'gnuplot')
 self.hwndText = windll.user32.FindWindowExA(self.hwndParent, None, 'wgnuplot_text', None)



 #
 # __del__
 #
 def __del__(self):
 windll.kernel32.CloseHandle(self.hProcess);
 windll.kernel32.CloseHandle(self.hThread);
 windll.user32.PostMessageA(self.hwndParent, WM_CLOSE, 0, 0)


 #
 # launch
 #
 def launch(self, path_to_exe):
 startupinfo = STARTUPINFO()
 process_information = PROCESS_INFORMATION()

 startupinfo.dwFlags = STARTF_USESHOWWINDOW
 startupinfo.wShowWindow = SW_HIDE

 if windll.kernel32.CreateProcessA(path_to_exe, None, None, None, False, 0, None, None, byref(startupinfo), byref(process_information)):
 self.hProcess = process_information.hProcess
 self.hThread = process_information.hThread
 else:
 print "Error: Create Process - Error code: ", windll.kernel32.GetLastError()
 sys.exit(1)



 #
 # execute
 #
 def execute(self, script, file_path):
 # make sure file doesn't exist
 try: os.unlink(file_path)
 except: pass

 # send script to gnuplot window
 for c in script: windll.user32.PostMessageA(self.hwndText, WM_CHAR, ord(c), 1L)

 # wait till gnuplot generates the chart
 while( not (os.path.exists(file_path) and (os.path.getsize(file_path) &gt; 0))): time.sleep(0.01)
</code></pre>

blocks|key|318134|text|我有点晚了，但由于我花了一些时间才让它工作，也许值得写个便条。这些程序在Windows上使用Python+3.3.2。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|318135|注意，字节是到处使用的，而不是字符串(例如b"plot+x"，而不是简单的"plot+x")，但如果有问题，只需执行以下操作：|318136|"plot+x".encode("ascii")|code-block|syntax|javascript|318137|第一个解决方案:使用通信来发送所有内容，并在完成时关闭。请不要忘记暂停，否则窗口会立即关闭。但是，如果使用gnuplot在文件中存储图像，这不是问题。|318138|from+subprocess+import+*
path+=+"C:\\app\\gnuplot\\bin\\gnuplot"
p+=+Popen([path],+stdin=PIPE,+stdout=PIPE)
p.communicate(b"splot+x*y\npause+4\n")|318139|第二种解决方案:使用stdin.write(...)一个接一个地发送命令。但是，不要忘记flush！(这是我一开始没有弄清楚的)，并使用terminate关闭连接，并在作业完成时使用gnuplot。|offset|length|style|BOLD|318140|from+subprocess+import+*
path+=+"C:\\app\\gnuplot\\bin\\gnuplot"
p+=+Popen([path],+stdin=PIPE,+stdout=PIPE)

p.stdin.write(b"splot+x*y\n")
p.stdin.flush()
...
p.stdin.write(b"plot+x,x*x\n")
p.stdin.flush()
...
p.terminate()|318141|entityMap^0|0|0|0|0|0|14|9|0|0^^$0|@$1|2|3|4|5|6|7|W|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|X|8|@]|9|@]|A|$]]|$1|D|3|E|5|F|7|Y|8|@]|9|@]|A|$G|H]]|$1|I|3|J|5|6|7|Z|8|@]|9|@]|A|$]]|$1|K|3|L|5|F|7|10|8|@]|9|@]|A|$G|H]]|$1|M|3|N|5|6|7|11|8|@$O|12|P|13|Q|R]]|9|@]|A|$]]|$1|S|3|T|5|F|7|14|8|@]|9|@]|A|$G|H]]|$1|U|3|-4|5|6|7|15|8|@]|9|@]|A|$]]]|V|$]]

I am a bit late, but since it took me some time to make it work, maybe it's worth putting a note. The programs are working with Python 3.3.2 on Windows.

Notice that bytes are used everywhere, not strings (e.g. b"plot x", not simply "plot x"), but in case it's a problem, simply do something like:

<pre><code>"plot x".encode("ascii")
</code></pre>

First solution: use communicate to send everything, and close when it's done. One must not forget pause, or the window is closed at once. However, it's not a problem if gnuplot is used to store images in files.

<pre><code>from subprocess import *
path = "C:\\app\\gnuplot\\bin\\gnuplot"
p = Popen([path], stdin=PIPE, stdout=PIPE)
p.communicate(b"splot x*y\npause 4\n")
</code></pre>

Second solution: send commands one after another, using stdin.write(...). But, don't forget flush! (this is what I didn't get right at first) And use terminate to close the connection and gnuplot when the job is done.

<pre><code>from subprocess import *
path = "C:\\app\\gnuplot\\bin\\gnuplot"
p = Popen([path], stdin=PIPE, stdout=PIPE)

p.stdin.write(b"splot x*y\n")
p.stdin.flush()
...
p.stdin.write(b"plot x,x*x\n")
p.stdin.flush()
...
p.terminate()
</code></pre>

blocks|key|318171|text|下面是另一个示例，它扩展了前面的一些答案。这个解决方案需要Gnuplot5.1，因为它使用了数据块。有关datablock的更多信息，请在gnuplot中执行help+datablocks。以前的一些方法的问题是，plot+'-'会立即使用紧跟在plot命令之后的数据。不能在后续的绘图命令中重复使用相同的数据。数据块可以用来缓解这个问题。使用数据块，我们可以模拟多个数据文件。例如，您可能想要使用来自两个数据文件的数据绘制图形，例如plot+"myData.dat"+using+1:2+with+linespoints,+''+using+1:3+with+linespoints,+"myData2.dat"+using+1:2+with+linespoints。我们可以直接将这些数据提供给gnuplot，而不需要创建实际的数据文件。|type|unstyled|depth|inlineStyleRanges|offset|length|style|BOLD|CODE|entityRanges|data|318172|import+sys,+subprocess
from+os+import+linesep+as+nl
from+subprocess+import+Popen,+PIPE


def+gnuplot(commands,+data):++++++++++++++++++++++++++++++++++++++++++++++++++++
++"""+drive+gnuplot,+expects+lists,+returns+stdout+as+string+"""++
++script=+nl.join(data)%2Bnl.join(commands)%2Bnl
++print+script
++args+=+["gnuplot",+"-p"]
++p+=+Popen(args,+shell=False,+stdin=PIPE)+++++++++++++++++++++++
++return+p.communicate(script)[0]++

def+buildGraph():
++commands+=+[\
++++++"set+datafile+separator+','",\
++++++"plot+'$data1'+using+1:2+with+linespoints,+''+using+1:3+with+linespoints,+'$data2'+using+1:2+with+linespoints",\
++++++]
++data+=+[\
++++++"$data1+<<+EOD",\
++++++"1,30,12",\
++++++"2,40,15",\
++++++"3,35,20",\
++++++"4,60,21",\
++++++"5,50,30",\
++++++"EOD",\
++++++"$data2+<<+EOD",\
++++++"1,20",\
++++++"2,40",\
++++++"3,40",\
++++++"4,50",\
++++++"5,60",\
++++++"EOD",\
++++++]

++return+(commands,+data)++


def+main(args):
++(commands,+data)+=+buildGraph()
++print+gnuplot(commands,+data)


if+__name__+==+"__main__":
+++main(sys.argv[1:])|code-block|syntax|javascript|318173|这种方法比plot+'-'更通用，因为它更容易多次重用相同的数据，包括在相同的plot命令上：https://stackoverflow.com/a/33064402/895245请注意，这种方法要求在使用plot命令之前将数据提供给gnuplot！|318174|此外，我没有像@ppetraki那样使用IOString，因为显然这比简单的列表连接器慢：https://waymoot.org/home/python_string/|318175|entityMap|0|LINK|mutability|MUTABLE|url|https://stackoverflow.com/a/33064402/895245|1|https://waymoot.org/home/python_string/^0|R|C|27|F|2Z|8|61|39|0|0|5|8|1B|17|0|0|19|13|1|0^^$0|@$1|2|3|4|5|6|7|Z|8|@$9|10|A|11|B|C]|$9|12|A|13|B|D]|$9|14|A|15|B|D]|$9|16|A|17|B|D]]|E|@]|F|$]]|$1|G|3|H|5|I|7|18|8|@]|E|@]|F|$J|K]]|$1|L|3|M|5|6|7|19|8|@$9|1A|A|1B|B|D]]|E|@$9|1C|A|1D|1|1E]]|F|$]]|$1|N|3|O|5|6|7|1F|8|@]|E|@$9|1G|A|1H|1|1I]]|F|$]]|$1|P|3|-4|5|6|7|1J|8|@]|E|@]|F|$]]]|Q|$R|$5|S|T|U|F|$V|W]]|X|$5|S|T|U|F|$V|Y]]]]

Here is another example which extends some of the previous answers. This solution requires Gnuplot 5.1 because it uses datablocks. For more information on datablocks, execute <code>help datablocks</code> in gnuplot.
The problem with some of the previous approaches is that <code>plot '-'</code> instantly consumes the data that immediately follows the plot command. It is not possible to reuse the same data in a subsequent plot command. datablocks can be used to alleviate this issue. Using datablocks we can mimic multiple datafiles. For instance, you may want to plot a graph using data from two data files, e.g. <code>plot "myData.dat" using 1:2 with linespoints, '' using 1:3 with linespoints, "myData2.dat" using 1:2 with linespoints</code>. We could feed this data directly to gnuplot without the need to create actual data files.

<pre><code>import sys, subprocess
from os import linesep as nl
from subprocess import Popen, PIPE


def gnuplot(commands, data): 
 """ drive gnuplot, expects lists, returns stdout as string """ 
 script= nl.join(data)+nl.join(commands)+nl
 print script
 args = ["gnuplot", "-p"]
 p = Popen(args, shell=False, stdin=PIPE) 
 return p.communicate(script)[0] 

def buildGraph():
 commands = [\
 "set datafile separator ','",\
 "plot '$data1' using 1:2 with linespoints, '' using 1:3 with linespoints, '$data2' using 1:2 with linespoints",\
 ]
 data = [\
 "$data1 &lt;&lt; EOD",\
 "1,30,12",\
 "2,40,15",\
 "3,35,20",\
 "4,60,21",\
 "5,50,30",\
 "EOD",\
 "$data2 &lt;&lt; EOD",\
 "1,20",\
 "2,40",\
 "3,40",\
 "4,50",\
 "5,60",\
 "EOD",\
 ]

 return (commands, data) 


def main(args):
 (commands, data) = buildGraph()
 print gnuplot(commands, data)


if __name__ == "__main__":
 main(sys.argv[1:])
</code></pre>

This method is a bit more versatile than <code>plot '-'</code> as it makes it easier to reuse the same data multiple times, including on the same plot command: <a href="https://stackoverflow.com/a/33064402/895245">https://stackoverflow.com/a/33064402/895245</a>
Notice that this approach requires that the data is fed to gnuplot before the plot commands!

Also, I did not use IOString as @ppetraki did, since apparently this is slower than a simple list joiner: <a href="https://waymoot.org/home/python_string/" rel="nofollow noreferrer">https://waymoot.org/home/python_string/</a>

I've a python script that after some computing will generate two data files formatted as gnuplot input.

How do I 'call' gnuplot from python ?

I want to send the following python string as input to gnuplot:

<pre><code>"plot '%s' with lines, '%s' with points;" % (eout,nout)
</code></pre>

where 'eout' and 'nout' are the two filenames.

PS:
I prefer not to use additional python modules (eg. gnuplot-py), only the standard API.

Thank You

Calling gnuplot from python

我有一个python脚本，经过一些计算后，它将生成两个格式化为gnuplot输入的数据文件。如何从python中“调用”gnuplot？我想将以下python字符串作为输入发送到gnuplot："plot '%s' with lines, '%s' with points;" % (eout,nout)其中'eout‘和'nout’是两个文件名。PS:我更喜欢不使用额外的python模块(例如.g

问从python调用gnuplot
EN

回答 8

Stack Overflow用户

Stack Overflow用户

Stack Overflow用户

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问从python调用gnuplotEN

回答 8

Stack Overflow用户

Stack Overflow用户

Stack Overflow用户

社区

活动

资源

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问从python调用gnuplot
EN