I have a problem with plotting in matplotlib. I need to plot a wide figure which represents an allocation of radio resources in time. But when a time period is big, the plot shrinks and I need to zoom it to see what is in particular fragment. What I want, is to plot the data without scaling ("real size") and to use the scrollbars to scroll the plot horizontally (in time). Is that possible?
3
- 1You may produce an image of your plot of arbitrary size and then use your favorite program to view it. I don't think any of the matplotlib backends supports scrollbars -- of course, you may just zoom in and use the hand tool to scroll...David Zwicker– David Zwicker2012-02-13 10:58:35 +00:00Commented Feb 13, 2012 at 10:58
- 1Okay, so I assume, that I need to use: fig.set_size_inches(XX,YY) fig.savefig('filename') And that seems to work, but is there any way to set the saved figure size as 1:1 size? Because it can be difficult to determine which size should I use for the next plot, and I will need to chenge script every time (or pass additional arg)?Kokos– Kokos2012-02-13 11:14:14 +00:00Commented Feb 13, 2012 at 11:14
- 1What do you mean by 1:1 size? What units are you using to specify your plot? And how shall those units relate to the pixels displayed by your screen?David Zwicker– David Zwicker2012-02-13 12:00:11 +00:00Commented Feb 13, 2012 at 12:00
Add a comment |
1 Answer
Here's an example from a scipy cookbook:
When plotting a very long sequence in a matplotlib canvas embedded in a wxPython application, it sometimes is useful to be able to display a portion of the sequence without resorting to a scrollable window so that both axes remain visible. Here is how to do so:
from numpy import arange, sin, pi, float, size import matplotlib matplotlib.use('WXAgg') from matplotlib.backends.backend_wxagg import FigureCanvasWxAgg from matplotlib.figure import Figure import wx class MyFrame(wx.Frame): def __init__(self, parent, id): wx.Frame.__init__(self,parent, id, 'scrollable plot', style=wx.DEFAULT_FRAME_STYLE ^ wx.RESIZE_BORDER, size=(800, 400)) self.panel = wx.Panel(self, -1) self.fig = Figure((5, 4), 75) self.canvas = FigureCanvasWxAgg(self.panel, -1, self.fig) self.scroll_range = 400 self.canvas.SetScrollbar(wx.HORIZONTAL, 0, 5, self.scroll_range) sizer = wx.BoxSizer(wx.VERTICAL) sizer.Add(self.canvas, -1, wx.EXPAND) self.panel.SetSizer(sizer) self.panel.Fit() self.init_data() self.init_plot() self.canvas.Bind(wx.EVT_SCROLLWIN, self.OnScrollEvt) def init_data(self): # Generate some data to plot: self.dt = 0.01 self.t = arange(0,5,self.dt) self.x = sin(2*pi*self.t) # Extents of data sequence: self.i_min = 0 self.i_max = len(self.t) # Size of plot window: self.i_window = 100 # Indices of data interval to be plotted: self.i_start = 0 self.i_end = self.i_start + self.i_window def init_plot(self): self.axes = self.fig.add_subplot(111) self.plot_data = \ self.axes.plot(self.t[self.i_start:self.i_end], self.x[self.i_start:self.i_end])[0] def draw_plot(self): # Update data in plot: self.plot_data.set_xdata(self.t[self.i_start:self.i_end]) self.plot_data.set_ydata(self.x[self.i_start:self.i_end]) # Adjust plot limits: self.axes.set_xlim((min(self.t[self.i_start:self.i_end]), max(self.t[self.i_start:self.i_end]))) self.axes.set_ylim((min(self.x[self.i_start:self.i_end]), max(self.x[self.i_start:self.i_end]))) # Redraw: self.canvas.draw() def OnScrollEvt(self, event): # Update the indices of the plot: self.i_start = self.i_min + event.GetPosition() self.i_end = self.i_min + self.i_window + event.GetPosition() self.draw_plot() class MyApp(wx.App): def OnInit(self): self.frame = MyFrame(parent=None,id=-1) self.frame.Show() self.SetTopWindow(self.frame) return True if __name__ == '__main__': app = MyApp() app.MainLoop() - The example needs
wxPython,numpyandmatplotlib;pip install numpy matplotlib wxPythonto install them.
Source: https://scipy-cookbook.readthedocs.io/items/Matplotlib_ScrollingPlot.html
6 Comments
Yann
Can you give more detail than a link?
Yann
I agree, your answer is spot on... I guess I was looking for a quick summary of how it does it, in case the scipy site was down. But looking at the code, it seems like saying more than "using a wxPython application" would involve a word for word explaining what the code does. +1 for the right answer, I was just sharing my impressions of the answer.
ev-br
OK, you're right in that it's worth saying a couple of words rather than just a link. An explanation of what a code does is indeed not worth it, but a one-sentence preamble is.
Kokos
Hi! Sorry for my long absence. But now I'm back ;) This link is exactly what I was looking for, but still do not cover all of my needs... but this is a topic for later. I have tried to combine this code, with my plotting code - see edited question.
Kanmani
the link doesn't work. They seemed to have removed the scrolling plot example. I searched the cookbook and nothing came up. Do you know any other sources for this problem ? Please help.
|