File:Frequency response of 3-term boxcar filter.svg

此SVG文件的PNG预览的大小：416 × 400像素。其他分辨率：250 × 240像素 | 499 × 480像素 | 799 × 768像素 | 1,065 × 1,024像素 | 2,130 × 2,048像素。

原始文件 ‎（SVG文件，尺寸为416 × 400像素，文件大小：20 KB）

本文件并非来自中文维基百科，而是来自维基共享资源。请参阅维基共享资源上的详细描述、讨论页、页面历史、日志。

维基共享资源是一个储存自由版权作品的项目。您可以向此项目作出贡献。

摘要

描述

English: Frequency response (discrete time Fourier transform) of a 3-tap boxcar filter. To produce these plots, samples of the DTFT are obtained by computing the DFT of the zero-filled impulse response. Alternatively, one can directly sample the closed form of the DTFT:

\scriptstyle H(\omega )={\frac {1}{3}}\left(1+e^{-j\omega }+e^{-j2\omega }\right),

and plot the magnitude and angle of the samples. (See File:MA2Amp_C.svg)
Note: At the two frequencies where the magnitude goes to zero, there are phase discontinuities of π radians. They are not artifacts of phase wrapping. Rather, they indicate a sign inversion... multiplication by -1. Since that is equivalent to an amplitude modification, it does not affect the filter's property of phase linearity. An illustration of that can be viewed at File:Amplitude & phase vs frequency for 3-term boxcar filter.svg.

日期

2017年8月22日

来源

自己的作品

作者

original: Bob K
vector version: Krishnavedala

授权
(二次使用本文件)

我，本作品著作权人，特此采用以下许可协议发表本作品：

	本作品采用知识共享 CC0 1.0 通用公有领域贡献许可协议授权。
	采用本宣告发表本作品的人，已在法律允许的范围内，通过在全世界放弃其对本作品拥有的著作权法规定的所有权利（包括所有相关权利），将本作品贡献至公有领域。您可以复制、修改、传播和表演本作品，将其用于商业目的，无需要求授权。 CC0falsefalse

其他版本

本檔案衍生自： Frequency response of 3-term boxcar filter.gif

SVG开发

该SVG的源代码为有效代码。

本矢量图由Krishnavedala使用Python创作。

Python source

click to expand

This script is a translation of the original Octave script into Python, for the purpose of generating an SVG file to replace the GIF version.

import scipy
from scipy import signal
import numpy as np
from matplotlib import pyplot as plt
plt.rcParams['svg.fonttype'] = 'none'

N = 256
h = np.array([1., 1., 1.]) / 3
H = scipy.fftpack.shift(scipy.fft(h, N), np.pi)
w = np.linspace(-N/2, N/2-1, num=N) * 2 * np.pi / N

fig = plt.figure(figsize=[5,5])
plt.subplot(211)
plt.plot(w, np.abs(H), 'blue')
plt.grid(True)
plt.ylabel('Magnitude')
plt.xlim([-np.pi,np.pi])
plt.xticks([-np.pi, -2*np.pi/3,0,2*np.pi/3,np.pi], [])
plt.subplot(212)
plt.plot(w, np.angle(H), 'blue')
plt.grid(True)
plt.ylabel('Phase (radians)')
plt.xlabel('$\\longleftarrow$ Frequency ($\\omega$) (radians/sample) $\\longrightarrow$')
plt.xticks([-np.pi, -2*np.pi/3,0,2*np.pi/3,np.pi], ['-$\pi$','-2$\pi$/3','0','2$\pi$/3','$\pi$'])
plt.xlim([-np.pi,np.pi])
plt.yticks([-np.pi, -2,-1,0,1,2,np.pi], ['-$\pi$','-2','-1','0','1','2','$\pi$'])
plt.ylim([-np.pi,np.pi])
plt.subplots_adjust(hspace=0.1)
plt.savefig('Frequency response of 3-term boxcar filter.svg', bbox_inches='tight', transparent=True)

Octave/gnuplot source

click to expand

This script was derived from the original in order to address some GNUplot bugs: a missing title and two missing axis labels.  And to add an Octave print function, which creates an SVG file.  Alternatively, the gnuplot screen image has an export function that produces an SVG file, but the π characters aren't as professional-looking.  I think the resultant quality produced by this script is now better than the file produced by the Python script.
graphics_toolkit gnuplot
clear all; close all; clc
  
hfig = figure("position",[100 100 509 509]);

x1 = .12;               % left margin for name of Y-variable
x2 = .02;               % right margin
y1 = .10;               % bottom margin for ticks
y2 = .08;               % top margin for title
dy = .08;               % vertical space between rows

width = 1-x1-x2;
height= (1-y1-y2-dy)/2; % space allocated for each of 2 rows

x_origin = x1;
y_origin = 1;           % start at top of graph area
%=======================================================
  N= 256;
  h = [1 1 1]/3;              % impulse response
  H = fftshift(fft(h,N));     % samples of DTFT
  abscissa = (-N/2:N/2-1)*2*pi/N;  % normalized frequency
%=======================================================
  y_origin = y_origin -y2 -height;        % position of top row
  subplot("position",[x_origin y_origin width height])
  plot(abscissa, abs(H), "linewidth",2);
  xlim([-pi pi])
  ylim([0 1.2])
  set(gca, "XTick", [-pi -2*pi/3 0 2*pi/3 pi])
  set(gca, "YTick", [0 .2 .4 .6 .8 1])
  grid("on")

  ylabel("Magnitude")
% set(gca, "ticklabelinterpreter", "tex")	% tex is the default
  set(gca, "XTickLabel", ['-\pi'; '-2\pi/3'; '0'; '2\pi/3'; '\pi';])
  set(gca, "YTickLabel", ['0'; '.2'; '.4'; '.6'; '.8'; '1';])
  title("Frequency response of 3-term boxcar filter", "fontsize", 12)
%=======================================================
  y_origin = y_origin -dy -height;
  subplot("position",[x_origin y_origin width height])
  plot(abscissa, angle(H), "linewidth",2);
  xlim([-pi pi])
  ylim([-pi pi])
  set(gca, "XTick", [-pi -2*pi/3 0 2*pi/3 pi])
  set(gca, "YTick", [-pi -2 -1 0 1 2 pi])
  grid("on")

  xlabel('\leftarrow Frequency (\omega) (radians/sample) \rightarrow')
  ylabel("Phase (radians)")
% set(gca, "ticklabelinterpreter", "tex")	% tex is the default
  set(gca, "XTickLabel", ['-\pi'; '-2\pi/3'; '0'; '2\pi/3'; '\pi';])
  set(gca, "YTickLabel", ['-\pi'; '-2'; '-1'; '0'; '1'; '2'; '\pi';])

% The print function results in nicer-looking "pi" symbols
% than the export function on the GNUPlot figure toolbar.
print(hfig,"-dsvg", "-S509,509","-color", ...
'C:\Users\BobK\Frequency response of 3-term boxcar filter.svg')

文件历史

点击某个日期/时间查看对应时刻的文件。

	日期/时间	大小	用户	备注
当前	2020年10月1日 (四) 13:07	416 × 400（20 KB）	Krishnavedala	Text-to-graph aspect ratio renders poorly in thumbnails with text unreadable.
	2019年7月3日 (三) 01:36	512 × 512（37 KB）	Bob K	Increase graph "linewidth" to 2.
	2019年7月2日 (二) 22:11	512 × 512（37 KB）	Bob K	Enlarge image. Add title. Improve rendering of "pi" symbols.
	2017年8月22日 (二) 15:35	416 × 400（20 KB）	Krishnavedala	User created page with UploadWizard