[dygraphs.git] / docs / index.html

<html>
  <head>
    <title>dygraphs JavaScript Library</title>
    <!--[if IE]>
    <script type="text/javascript" src="excanvas.js"></script>
    <![endif]-->
    <script type="text/javascript" src="dygraph-combined.js"></script>
    <style type="text/css">
      .thinborder {
        border-width: 1px;
        border-spacing: 0px;
        border-style: solid;
        border-color: black;
        border-collapse: collapse;
      }
      .thinborder td, .thinborder th {
        border-width: 1px;
        padding: 5px;
        border-style: solid;
        border-color: black;
      }
    </style>
  </head>
<body>
  <center>
    <p><span style="font-size:28pt;">dygraphs JavaScript Library</span><br/>
    <a href="http://code.google.com/p/dygraphs/">code.google.com/p/dygraphs</a></p>
  </center>

<p>The dygraphs JavaScript library produces produces interactive, zoomable charts of time series.</p>

<h3>Features</h3>
<ul>
  <li>Plots time series without using an external server or Flash</li>
  <li>Supports multiple data series</li>
  <li>Supports error bands around data series</li>
  <li>Displays values on mouseover</li>
  <li>Interactive zoom</li>
  <li>Adjustable averaging period</li>
  <li>Customizable click-through actions</li>
  <li>Compatible with the Google Visualization API</li>
</ul>

<h3>Caveats</h3>
<ul>
  <li>Requires Firefox 1.5+ or Safari/WebKit 1.3+.</li>
  <li>Internet Explorer is poorly supported.</li>
</ul>

<h2>Demo</h2>
<font size=-1>(Mouse over to highlight individual values. Click and drag to zoom. Double-click to zoom out.)</font><br/>
<table><tr><td>
<div id="demodiv" style="width:480px; height:320px;"></div>
</td><td valign=top>
<div id="status" style="width:200px; font-size:0.8em; padding-top:5px;"></div>
</td>
</tr></table>
<script type="text/javascript">
  g = new DateGraph(
          document.getElementById("demodiv"),
          function() {
            var zp = function(x) { if (x < 10) return "0"+x; else return x; };
            var r = "date,parabola,line,another line,sine wave\n";
            for (var i=1; i<=31; i++) {
            r += "200610" + zp(i);
            r += "," + 10*(i*(31-i));
            r += "," + 10*(8*i);
            r += "," + 10*(250 - 8*i);
            r += "," + 10*(125 + 125 * Math.sin(0.3*i));
            r += "\n";
            }
            return r;
          },
          null,
          {
            rollPeriod: 1,
            labelsDiv: document.getElementById('status'),
            labelsSeparateLines: true,
            labelsKMB: true,
            colors: ["hsl(180,60,50)",
                     "rgb(255,100,100)",
                     "#00DD55",
                     "rgba(50,50,200,0.4)"],
            padding: {left: 40, right: 30, top: 15, bottom: 15},
          }
      );
</script>

<h2>Usage</h2>

<p>The dygraphs library depends on two other JS libraries: <a href="http://www.mochikit.com/">MochiKit</a> and <a href="http://www.liquidx.net/plotkit/">PlotKit</a>. Rather than tracking down copies of these libraries, I recommend using a packed version of dygraphs that combines all three libraries into a single JS file. Either grab this file from dygraph project's <a href="http://code.google.com/p/dygraphs/downloads/list">downloads</a> page or create it yourself by <a href="http://code.google.com/p/dygraphs/source/checkout">checking out</a> a copy of the code and running:

<pre>./generate-combined.sh</pre>

<p>The combined JS file is now in <code>dygraph-combined.js</code>. Here's a basic example to get things started:</p>

<table>
  <tr><th>HTML</th>
      <td rowspan=2><img src=arrow.gif /></td>
      <th>Output</th></tr>
<tr>
<td valign=top><pre>
&lt;html&gt;
&lt;head&gt;
&lt;script type="text/javascript" src="combined.js"&gt;&lt;/script&gt;
&lt;/head&gt;
&lt;body&gt;
&lt;div id="graphdiv"&gt;&lt;/div&gt;
&lt;script type="text/javascript"&gt;
  g = new Dygraph(
        document.getElementById("graphdiv"),  // containing div
        "Date,Temperature\n" +                // CSV or path to a CSV file.
        "20080507,75\n" +
        "20080508,70\n" +
        "20080509,80\n",
      );
&lt;/script&gt;
&lt;/body&gt;
&lt;/html&gt;
</pre>
</td><td valign=top>
  <div id="graphdiv"></div>
  <script type="text/javascript">
    g = new Dygraph(
        document.getElementById("graphdiv"),  // containing div
        "Date,Temperature\n" +                // CSV or path to a CSV file.
        "20080507,75\n" +
        "20080508,70\n" +
        "20080509,80\n"
      );
  </script>
</td></tr></table>

<p>In order to keep this example self-contained, the second parameter is a function that returns CSV data. These lines <i>must</i> begin with a date in the form <i>YYYYMMDD</i>. In most applications, it makes more sense to include a CSV file instead. If the second parameter to the constructor is a string, it will be interpreted as the path to a CSV file. The Dygraph will perform an XMLHttpRequest to retrieve this file and display the data when it becomes available. Make sure your CSV file is readable and serving from a place that understands XMLHttpRequest's! In particular, you cannot specify a CSV file using <code>"file:///"</code>. Here's an example: (data from <a href="http://www.wunderground.com/history/airport/KNUQ/2007/1/1/CustomHistory.html?dayend=31&monthend=12&yearend=2007&req_city=NA&req_state=NA&req_statename=NA">Weather Underground</a>)</p>

<table>
  <tr><th>HTML</th>
      <td rowspan=2><img src=arrow.gif /></td>
      <th>Output</th></tr>
<tr>
<td valign=top><pre>
&lt;html&gt;
&lt;head&gt;
&lt;script type="text/javascript" src="combined.js"&gt;&lt;/script&gt;
&lt;/head&gt;
&lt;body&gt;
&lt;div id="graphdiv" style="width:600px; height:300px;"&gt;&lt;/div&gt;
&lt;script type="text/javascript"&gt;
  g = new Dygraph(
        document.getElementById("graphdiv"),
        "temperatures.csv",  // path to CSV file
        {}                   // additional options
      );
&lt;/script&gt;
&lt;/body&gt;
&lt;/html&gt;
</pre>
</td><td valign=top>
  <div id="graphdiv2" style="width:600px; height:300px;"></div>
  <script type="text/javascript">
    g2 = new Dygraph(
          document.getElementById("graphdiv2"),
          "temperatures.csv", {}
        );
  </script>
</td></tr></table>

<p>Click <a href="temperatures.csv">here</a> to view the <code>temperatures.csv</code> file. There are a few things to note here:</p>

<ul>
  <li>The Dygraph sent off an XHR to get the temperatures.csv file.</li>
  <li>The labels were taken from the first line of <code>temperatures.csv</code>, which is <code>Date,High,Low</code>.</li>
  <li>The Dygraph automatically chose two different, easily-distinguishable colors for the two data series.</li>
  <li>The labels on the x-axis have switched from days to months. If you zoom in, they'll switch to weeks and then days.</li>
  <li>Some heuristics are used to determine a good vertical range for the data. The idea is to make all the data visible and have human-friendly values on the axis (i.e. 200 instead of 193.4). Generally this works well.</li>
  <li>The data is very spiky. A moving average would be easier to interpret.</li>
</ul>

<p>This problem can be fixed by specifying the appropriate options in the "additional options" parameter to the Dygraph constructor. To set the number of days for a moving average, use the <b>rollPeriod</b> option. Here's how it's done:</p>

<table>
  <tr><th>HTML</th>
      <td rowspan=2><img src=arrow.gif /></td>
      <th>Output</th></tr>
<tr>
<td valign=top><pre>
&lt;html&gt;
&lt;head&gt;
&lt;script type="text/javascript" src="combined.js"&gt;&lt;/script&gt;
&lt;/head&gt;
&lt;body&gt;
&lt;div id="graphdiv" style="width:600px; height:300px;"&gt;&lt;/div&gt;
&lt;script type="text/javascript"&gt;
  g = new Dygraph(
        document.getElementById("graphdiv"),
        "temperatures.csv",
        { rollPeriod: 7,
          showRoller: true,
        }
      );
&lt;/script&gt;
&lt;/body&gt;
&lt;/html&gt;
</pre>
</td><td valign=top>
  <div id="graphdiv3" style="width:600px; height:300px;"></div>
  <script type="text/javascript">
    g3 = new Dygraph(
          document.getElementById("graphdiv3"),
          "temperatures.csv",
          { rollPeriod: 7,
            showRoller: true,
          }
        );
  </script>
</td></tr></table>

<p>A rolling average can be set using the text box in the lower left-hand corner of the graph (the showRoller attribute is what makes this appear).</p>

<h2>Error Bars</h2>
<p>Another significant feature of the dygraphs library is the ability to display error bars around data series. One standard deviation must be specified for each data point. A +/-<i>n</i> sigma band will be drawn around the data series at that point. If a moving average is being displayed, dygraphs will compute the standard deviation of the average at each point. (i.e. <i>&sigma;</i> = sqrt((<i>&sigma;_1</i>^2 + <i>&sigma;_2</i>^2 + ... + <i>&sigma;_n</i>^2)/<i>n</i>))</p>

<p>Here's a demonstration. There are two data series. One is <code>N(100,10)</code> with a standard deviation of 10 specified at each point. The other is <code>N(80,20)</code> with a standard deviation of 20 specified at each point. The CSV file was generated using Octave and can be viewed <a href="twonormals.csv">here</a>.</p>

<table>
  <tr><th>HTML</th>
      <td rowspan=2><img src=arrow.gif /></td>
      <th>Output</th></tr>
<tr>
<td valign=top><pre>
&lt;html&gt;
&lt;head&gt;
&lt;script type="text/javascript"
  src="combined.js"&gt;&lt;/script&gt;
&lt;/head&gt;
&lt;body&gt;
&lt;div id="graphdiv" 
 style="width:800px; height:400px;"
 &gt;&lt;/div&gt;
&lt;script type="text/javascript"&gt;
$ = document.getElementById;
g = new Dygraph(
  $("graphdiv"),
  "twonormals.csv",
  { rollPeriod: 7,
    showRoller: true,
    errorBars: true,
    valueRange: [50,125]
  }
);
&lt;/script&gt;
&lt;/body&gt;
&lt;/html&gt;
</pre>
</td><td valign=top>
  <div id="graphdiv4" style="width:800px; height:400px;"></div>
  <script type="text/javascript">
$ = document.getElementById;
new Dygraph(
  document.getElementById("graphdiv4"),
  "twonormals.csv",
  { rollPeriod: 14,
    showRoller: true,
    errorBars: true,
    valueRange: [50, 125]
  }
);
  </script>
</td></tr></table>

<p>Things to note here:</p>
<ul>
  <li>The <b>errorBars</b> option affects both the interpretation of the CSV file and the display of the graph. When <b>errorBars</b> is set to true, each line is interpreted as <i>YYYYMMDD</i>,<i>A</i>,<i>sigma_A</i>,<i>B</i>,<i>sigma_B</i>,...</li>
  <li>The first line of the CSV file doesn't mention the error columns. In this case, it's just "Date,Series1,Series2".</li>
  <li>The averaging visibly affects the error bars. This is most clear if you crank up the rolling period to something like 100 days. For the earliest dates, there won't be 100 data points to average so the signal will be noisier. The error bars get smaller like sqrt(N) going forward in time until there's a full 100 points to average.</li>
  <li>The error bars are partially transparent. This can be seen when they overlap one another.</li>
</ul>

<h2>One last demo</h2>

<p>This chart shows monthly closes of the Dow Jones Industrial Average, both in nominal and real (i.e. adjusted for inflation) dollars. The shaded areas show its monthly high and low. CPI values with a base from 1982-84 are used to adjust for inflation.</p>

<div id=dow_chart style="width:1000px; height:400px;"></div>
<script type="text/javascript">
  // From http://www.econstats.com/eqty/eq_d_mi_3.csv
  dow = new Dygraph(
    document.getElementById('dow_chart'),
    "dow.txt",
    {
      showRoller: true,
      customBars: true,
      labelsKMB: true,
      padding: {left:30, right:30, top:5, bottom:5}
    });
</script>
<!--

Here is a script to regenerate the Dow Jones plot:

# Get unadjusted DJIA data in a nice format:
curl -O http://www.econstats.com/eqty/eq_d_mi_3.csv
sed '1,17d' eq_d_mi_3.csv | cut -d, -f1,6 | perl -pe 's/(\d{4}-\d\d)-\d\d/$1/g' | perl -pe 's/, */\t/' | grep -v 'na' | perl -ne 'chomp; ($m,$v) = split/\t/; $close{$m} = $v; if ($low{$m} == 0 || $v < $low{$m}) { $low{$m}=$v } if ($v > $high{$m}) { $high{$m} = $v } END { for $x(sort keys %close) { print "$x\t$low{$x}\t$close{$x}\t$high{$x}\n" } } ' > monthly-djia.tsv

# Fetch and format the CPI data:
curl 'http://data.bls.gov/PDQ/servlet/SurveyOutputServlet?series_id=CUUR0000SA0&years_option=all_years&periods_option=all_periods&output_type=column&output_format=text&delimiter=comma' > cpi-u.txt
sed '1,/Series Id,Year,/d' cpi-u.txt | sed '/^$/,$d' | cut -d, -f2,3,4 | perl -ne 'print if /,M(0[0-9]|1[012]),/' | perl -pe 's/(\d{4}),M(\d{2}),/$1-$2\t/g' > cpi-u.tsv

# Merge:
join -t'       ' cpi-u.tsv monthly-djia.tsv > annotated-djia.tsv
perl -ne 'BEGIN{print "Month,Nominal,Real\n"} chomp; ($m,$cpi,$low,$close,$high) = split /\t/; $cpi /= 100.0; print "$m-15,$low;$close;$high,",($low/$cpi),";",($close/$cpi),";",($high/$cpi),"\n"' annotated-djia.tsv > dow.txt

-->


<h2>Other Options</h2>
<p>These are the options that can be passed in through the optional third parameter of the Dygraph constructor.</p>

<table class=thinborder width=1000>
  <tr><th>Name</th><th>Sample Value</th><th>Description</th></tr>
  <tr>
    <td><b>rollPeriod</b></td>
    <td><code>7</code></td>
    <td>Number of days over which to average data. Discussed extensively above.</td>
  </tr>
  
  <tr>
    <td><b>showRoller</b></td>
    <td><code>true</code></td>
    <td>Should the rolling average period text box be shown? Default is false.</td>
  </tr>
  
  <tr>
    <td><b>colors</b></td>
    <td><code>['red',&nbsp;'#00FF00']</code></td>
    <td>List of colors for the data series. These can be of the form "#AABBCC"
      or "rgb(255,100,200)" or "yellow", etc. If not specified, equally-spaced
      points around a color wheel are used.</td>
  </tr>
  
  <tr>
    <td><b>colorSaturation</b></td>
    <td><code>1.0</code></td>
    <td>If <b>colors</b> is not specified, saturation of the
      automatically-generated data series colors. (0.0-1.0, default:
      1.0)</td>
  </tr>
  
  <tr>
    <td><b>colorValue</b></td>
    <td><code>0.5</code></td>
    <td>If colors is not specified, value of the data series colors, as in
      hue/saturation/value. (0.0-1.0, default 0.5)</td>
  </tr>
  
  <tr>
    <td><b>clickCallback</b></td>
    <td><code>function(e,date){ alert(date); }</code></td>
    <td>A function to call when a data point is clicked. The function should take
      two arguments, the event object for the click and the date that was
      clicked. (default null)</td>
  </tr>

  <tr>
    <td><b>zoomCallback</b></td>
    <td><code>function(minDate,maxDate) {}</code></td>
    <td>A function to call when the zoom window is changed (either by zooming
    in or out). minDate and maxDate are millis since epoch.</td>
  </tr>
  
  <tr>
    <td><b>strokeWidth</b></td>
    <td><code>2.0</code></td>
    <td>Width of the data lines. This can be used to increase the contrast or
      some graphs. (default 1.0)</td>
  </tr>
  
  <tr>
    <td><b>dateWindow</b></td>
    <td><code>[Date.parse('2006-01-01'),<br/>
        (new&nbsp;Date()).valueOf()]</code></td>
    <td>Initially zoom in on a section of the graph. Is of the form [earliest,
      latest], where earliest/latest are millis since epoch.  By default, the
      full range of the input is shown.</td>
  </tr>

  <tr>
    <td><b>valueRange</b></td>
    <td><code>[10, 110]</code></td>
    <td>Explicitly set the vertical range of the graph to [low, high]. By
      default, some clever heuristics are used (see above).</td>
  </tr>
  
  <tr>
    <td><b>labelsSeparateLines</b></td>
    <td><code>true</code></td>
    <td>Put &lt;br/&gt; between lines in the label string. Often used in
      conjunction with <b>labelsDiv</b>. (default false)</td>
  </tr>
  
  <tr>
    <td><b>labelsDiv</b></td>
    <td><code>document.getElementById('foo')</code></td>
    <td>Show data labels in an external div, rather than on the graph. (default
      null)</td>
  </tr>
  
  <tr>
    <td><b>labelsKMB</b></td>
    <td><code>true</code></td>
    <td>Show K/M/B for thousands/millions/billions on y-axis (default
      false).</td>
  </tr>

  <tr>
    <td><b>labelsDivWidth</b></td>
    <td>250</td>
    <td>Width (in pixels) of the div which shows information on the
    currently-highlighted points.</td>
  </tr>

  <tr>
    <td><b>labelsDivStyles</b></td>
    <td>{}</td>
    <td>Additional styles to apply to the currently-highlighted points div. For
    example, { 'font-weigth': 'bold' } will make the labels bold.</td>
  </tr>
  
  <tr>
    <td><b>highlightCircleSize</b></td>
    <td><code>3</code></td>
    <td>Size (in pixels) of the dot drawn over highlighted points (default 3).</td>
  </tr>

  <tr>
    <td><b>pixelsPerXLabel</b>, <b>pixelsPerYLabel</b></td>
    <td>50</td>
    <td>Number of pixels to require between each x- and y-label. Larger values
    will yield a sparser axis with fewer ticks. Defaults: 60 (x-axis), 30
    (y-axis).</td>
  </tr>

  <tr>
    <td><b>xAxisLabelWidth</b>, <b>yAxisLabelWidth</b></td>
    <td>50</td>
    <td>Width (in pixels) of the x- and y-axis labels.</td>
  </tr>

  <tr>
    <td><b>axisLabelFontSize</b></td>
    <td>14</td>
    <td>Size of the font (in pixels) to use in the axis labels, both x- and
    y-axis.</td>
  </tr>

  <tr>
    <td><b>rightGap</b></td>
    <td>5</td>
    <td>Number of pixels to leave blank at the right edge of the Dygraph. This
    makes it easier to highlight the right-most data point.</td>
  </tr>
  
  <tr>
    <td><b>errorBars</b></td>
    <td><code>false</code></td>
    <td>Does the data contain standard deviations? Setting this to true alters
      the input format (see above). (default false)</td>
  </tr>

  <tr>
    <td><b>sigma</b></td>
    <td>2</td>
    <td>When errorBars is set, shade this many standard deviations above/below
    each point.</td>
  </tr>

  <tr>
    <td><b>fractions</b></td>
    <td>false</td>
    <td>When set, attempt to parse each cell in the CSV file as "a/b", where a
    and b are integers. The ratio will be plotted. This allows computation of
    Wilson confidence intervals (see below).</td>
  </tr>

  <tr>
    <td><b>wilsonInterval</b></td>
    <td>true</td>
    <td>Use in conjunction with the "fractions" option. Instead of plotting +/-
    N standard deviations, dygraphs will compute a Wilson confidence interval
    and plot that. This has more reasonable behavior for ratios close to 0 or
    1.</td>
  </tr>

  <tr>
    <td><b>customBars</b></td>
    <td>false</td>
    <td>When set, parse each CSV cell as "low;middle;high". Error bars will be
    drawn for each point between low and high, with the series itself going
    through middle.</td>
  </tr>
</table>

<p>Any options you specify also get passed on to PlotKit's <a href="http://media.liquidx.net/js/plotkit-doc/PlotKit.Renderer.html">Renderer</a> class. dygraphs will override some of these (e.g. strokeColor), but others may be useful. The <code>padding</code> property is an example of this.</p>

<h2>Common Gotchas</h2>
<p>Here are a few problems that I've frequently run into while using the
dygraphs library.</p>

<ul>
  <li>Make sure your CSV files are readable! If your graph isn't showing up,
  the XMLHttpRequest for the CSV file may be failing. You can determine whether
  this is the case using tools like <a
    href="http://www.getfirebug.com/">Firebug</a>.</li>

  <li>Make sure your CSV files are in the correct format. They must be of the
  form <code>YYYYMMDD,series1,series2,...</code>. And if you set the
  <code>errorBars</code> property, make sure you alternate data series and
  standard deviations.</li>

  <li>dygraphs are not happy when placed inside a <code>&lt;center&gt;</code>
  tag. This applies to the CSS <code>text-align</code> property as well. If you
  want to center a Dygraph, put it inside a table with "align=center"
  set.</li>

  <li>Don't set the <code>dateWindow</code> property to a date. It expects
  milliseconds since epoch, which can be obtained from a JavaScript Date
  object's valueOf method.</li>
</ul>

<p><font size=-1>Created May 9, 2008 by <a href=mailto:danvdk@gmail.com>Dan Vanderkam</a></font></p>

</body>
</html>