A graph of the Hilbert transform (red) of a square wave (blue).

File:Hilbert transform.svg是本檔案的向量版本。 如果品質不低，就應該優先使用該檔案，而非PNG檔案。 File:Hilbert transform.png → File:Hilbert transform.svg 更多資訊請參閱Help:SVG/zh。 其他語言 Alemannisch ∙ Bahasa Indonesia ∙ Bahasa Melayu ∙ British English ∙ català ∙ čeština ∙ dansk ∙ Deutsch ∙ eesti ∙ English ∙ español ∙ Esperanto ∙ euskara ∙ français ∙ Frysk ∙ galego ∙ hrvatski ∙ Ido ∙ italiano ∙ lietuvių ∙ magyar ∙ Nederlands ∙ norsk bokmål ∙ norsk nynorsk ∙ occitan ∙ Plattdüütsch ∙ polski ∙ português ∙ português do Brasil ∙ română ∙ Scots ∙ sicilianu ∙ slovenčina ∙ slovenščina ∙ suomi ∙ svenska ∙ Tiếng Việt ∙ Türkçe ∙ vèneto ∙ Ελληνικά ∙ беларуская (тарашкевіца) ∙ български ∙ македонски ∙ нохчийн ∙ русский ∙ српски / srpski ∙ татарча/tatarça ∙ українська ∙ ქართული ∙ հայերեն ∙ বাংলা ∙ தமிழ் ∙ മലയാളം ∙ ไทย ∙ 한국어 ∙ 日本語 ∙ 简体中文 ∙ 繁體中文 ∙ עברית ∙ العربية ∙ فارسی ∙ +/−

 

Instructions

See Wikipedia graph-making tips.

 

本PNG 點陣圖使用Gnuplot創作。

Plotted in Maxima with the following commands:

Heaviside step function:

u(t) := 1/2*(1+signum(t));

Rectangular function:

rect(t) := u(t + 1/2) - u(t - 1/2);

A train of rectangular pulses (or a square wave). This should be n=−∞..+∞, but I can't get Maxima to evaluate it, so I just chose n values that would give a close enough approximation for the visible area:

train(t) := sum(rect(t-2*n),n,-10,20);

The Hilbert transform of a rectangular pulse:

hrect(t) := log(abs((t+1/2)/(t-1/2)))/%PI;

A "train" of hrects. Again with the approximation, though this time it actually matters (rectangular function is 0 everywhere; this is not.):

htrain(t) := sum(hrect(t-2*n),n,-100,100);

The plot command:

plot2d([train(t)-0.5,htrain(t)],[t,0,6],[y,-1,1],[nticks,500], [gnuplot_preamble,"set grid; set ytics 1; set nokey"], [gnuplot_ps_term_command,"set term postscript enhanced landscape color solid lw 2 'Times-Roman' 20"], [gnuplot_term,ps], [gnuplot_out_file,"hilbert.ps"]);

This causes gnuplot to create a .ps file.

To convert the PostScript file to PNG:

1. Open it in the GIMP (make sure you have ghostscript installed! - Windows instructions)
   • Enter 500 under Resolution (it doesn't say "DPI" but I think that's what it means)
   • Uncheck Try bounding box (since the bounding box cuts off the edge, unfortunately. You can try with the bounding box first.)
   • Enter large values for Width and Height
   • Check Color
   • Check Strong anti-aliasing for both graphics and text
2. Crop off extra whitespace (Shift+C if you can't find it in the toolbox)
3. Possibly need to rotate it: Click Image → Transform → Rotate 90 degrees clockwise
4. Filters → Blur → Gaussian blur at 2.0 px (No need to blur if you use strong anti-aliasing during conversion. I see no significant difference between end results.)
5. Image → Scale Image...
   • Width and Height at 25%
   • Cubic interpolation
6. You can view at normal size if you want by pressing 1, Ctrl+E
7. Save as Hilbert transform.png

Original version of this plot was made in GNU Octave with the command:

imag(hilbert(x))

Created by User:Omegatron using gnuplot, possibly with post-processing in the GIMP (PNG) or Inkscape (SVG)

我，本作品的著作權持有者，決定用以下授權條款發佈本作品：

此檔案採用創用CC 姓名標示-相同方式分享 3.0 未在地化版本、2.5 通用版、2.0 通用版以及1.0 通用版授權條款。

您可以自由：

• 分享 – 複製、發佈和傳播本作品
• 重新修改 – 創作演繹作品

惟需遵照下列條件：

• 姓名標示 – 您必須指名出正確的製作者，和提供授權條款的連結，以及表示是否有對內容上做出變更。您可以用任何合理的方式來行動，但不得以任何方式表明授權條款是對您許可或是由您所使用。
• 相同方式分享 – 如果您利用本素材進行再混合、轉換或創作，您必須基於如同原先的相同或兼容的條款，來分布您的貢獻成品。

https://creativecommons.org/licenses/by-sa/3.0CC BY-SA 3.0 Creative Commons Attribution-Share Alike 3.0 truetrue

已授權您依據自由軟體基金會發行的無固定段落、封面文字和封底文字GNU自由文件授權條款1.2版或任意後續版本，對本檔進行複製、傳播和／或修改。該協議的副本列在GNU自由文件授權條款中。http://www.gnu.org/copyleft/fdl.htmlGFDLGNU Free Documentation Licensetruetrue

您可以選擇您需要的授權條款。

(In short, this means that you can copy and modify the image freely as long as you provide attribution; preferably in the form of a link back to this page.)