Notebook

Charts with XPlot¶

Charts can be rendered using Xplot.Plotly. We will cover some example on how to use XPlot in a notebook with the .NET Kernel.

First, import the XPlot.Plotly namespace:

In [ ]:

#r "nuget: XPlot.Plotly.Interactive, 4.0.6"

In [ ]:

using XPlot.Plotly;

The next cell sets up some helpers for data generation.

In [ ]:

var generator = new Random();

Rendering Scatter plots¶

One of the most commonly used type of chart to explore data set. Use the type Scatter.

In [ ]:

var openSeries = new Scatter
{
    name = "Open",
    x = new[] {1, 2, 3, 4},
    y = new[] {10, 15, 13, 17}
};

var closeSeries = new Scatter
{
    name = "Close",
    x = new[] { 2,3,4,5 },
    y = new[] { 16, 5, 11, 9 }
};

var chart = Chart.Plot(new[] {openSeries, closeSeries});
chart.WithTitle("Open vs Close");
display(chart);

Let's change it to be markers style, so more like a scatter plot.

In [ ]:

openSeries.mode = "markers";
closeSeries.mode = "markers";
chart = Chart.Plot(new [] {openSeries, closeSeries});
display(chart);

Scatter can also produce polar charts by setting the radial property r and angular proeprty t

In [ ]:

openSeries = new Scatter
{
    name = "Open",
    r = new[] {1, 2, 3, 4},
    t = new[] {45, 100, 150, 290}
};

closeSeries = new Scatter
{
    name = "Close",
    r = new[] { 2,3,4,5 },
    t = new[] { 16, 45, 118, 90 }
};

chart = Chart.Plot(new[] {openSeries, closeSeries});
chart.WithLayout(new Layout.Layout
            {
                orientation = -90
            });
display(chart);

Large scatter plots and performance¶

It is not uncommong to have scatter plots with a large dataset, it is a common scenario at the beginning of a data exploration process. Using the default svg based rendering will create performace issues as the dom will become very large. We can then use web-gl support to address the problem.

In [ ]:

#!time
var series = new Scattergl[10];

for(var i = 0; i < series.Length; i++){
    series[i] =  new Scattergl
    {
        name = $"Series {i}",
        mode = "markers",
        x = Enumerable.Range(0,100000).Select(_ => generator.Next(-200, 200) * 1000 * generator.Next(-2000, 2000)),
        y = Enumerable.Range(0,100000).Select(_ => generator.Next(-200, 200) * 1000 * generator.Next(-2000, 2000))
    };
}

chart = Chart.Plot(series);
chart.WithTitle("Large Dataset");
display(chart);

Can provide custom marker colour, size and colorscale to display even more information to the user.

In [ ]:

for(var i = 0; i < series.Length; i++){
    var sizes = Enumerable.Range(0,100).Select(_ => (generator.NextDouble() < 0.75) ? generator.Next(1, 5) : generator.Next(10, 15)).ToArray();
    var temperatures = sizes.Select(t => (t * 10) - 100);
    
    series[i].x = Enumerable.Range(0,100).Select(_ => generator.Next(-200, 200) * 1000 * generator.Next(-2000, 2000));
    series[i].y = Enumerable.Range(0,100).Select(_ => generator.Next(-200, 200) * 1000 * generator.Next(-2000, 2000));
    series[i].marker = new Marker{        
        size = sizes,
        color = temperatures,
        colorscale = "hot"    
    };
}

chart = Chart.Plot(series);
chart.WithTitle("Size and Colour");
display(chart);

Plotly pvoides some additional color scales to use.

In [ ]:

for(var i = 0; i < series.Length; i++) {
    series[i].marker.colorscale = "Viridis";
}

chart = Chart.Plot(series);
chart.WithTitle("Viridis scale");
display(chart);

for(var i = 0; i < series.Length; i++) {
    series[i].marker.colorscale = "Hot";
}

chart = Chart.Plot(series);
chart.WithTitle("Hot scale");
display(chart);

for(var i = 0; i < series.Length; i++) {
    series[i].marker.colorscale = "Jet";
}

chart = Chart.Plot(series);
chart.WithTitle("Jet scale");
display(chart);

Rendering Histograms¶

Let's have a look at using histograms, the next cell sets up some generators.

In [ ]:

var count = 20;
var dates = Enumerable.Range(0, count).Select(i => DateTime.Now.AddMinutes(generator.Next(i, i + 30))).ToArray();

Now let's define histogram traces:

In [ ]:

var openByTime = new Histogram
{
    x = dates,
    y = Enumerable.Range(0, count).Select(_ => generator.Next(0,200)),
    name = "Open"
};

var closeByTime = new Histogram
{
    x = dates,
    y = Enumerable.Range(0, count).Select(_ => generator.Next(0, 200)),
    name = "Close"
};
chart = Chart.Plot(new [] {openByTime, closeByTime});
display(chart);

The Histogram generator will automatically count the number of items per bin.

Setting histfunc to "sum" we can now add up all the values contained in each bin. Note that we are creatng bin using the x data point and we are using bydefault autobinx

In [ ]:

openByTime.histfunc = "sum";;
closeByTime.histfunc = "sum";
chart = Chart.Plot(new [] {openByTime, closeByTime});
display(chart);

Area chart and Polar Area chart¶

By populating hte property fill of a Scatter trace the chart will render as area chart.

Here is set to "tozeroy" which will create a fill zone underneath the line reachin to the 0 of the y axis.

In [ ]:

openSeries = new Scatter
{
    name = "Open",
    x = new[] {1, 2, 3, 4},
    y = new[] {10, 15, 13, 17},
    fill = "tozeroy",
    mode= "lines"
};

closeSeries = new Scatter
{
    name = "Close",
    x = new[] {1, 2, 3, 4},
    y = new[] {3, 5, 11, 9},
    fill = "tozeroy",
    mode= "lines"
};

chart = Chart.Plot(new[] {openSeries, closeSeries});
chart.WithTitle("Open vs Close");
display(chart);

With one fill set to "tonexty" the cahrt will fill the aread between traces.

In [ ]:

openSeries.fill = null;
closeSeries.fill = "tonexty";

chart = Chart.Plot(new[] {openSeries, closeSeries});
chart.WithTitle("Open vs Close");
display(chart);

Using Area traces we can generate radial area chart. In this example we are using cardinal points to xpress angular values. The array {"North", "N-E", "East", "S-E", "South", "S-W", "West", "N-W"} will be autoimatically translated to angular values.

In [ ]:

var areaTrace1 =
    new Area
    {
        r = new[] {77.5, 72.5, 70.0, 45.0, 22.5, 42.5, 40.0, 62.5},
        t = new[] {"North", "N-E", "East", "S-E", "South", "S-W", "West", "N-W"},
        name = "11-14 m/s",
        marker = new Marker
        {
            color = "rgb(106,81,163)"
        }
    };

var areaTrace2 =
    new Area
    {
        r = new  [] {57.49999999999999, 50.0, 45.0, 35.0, 20.0, 22.5, 37.5, 55.00000000000001},
        t = new [] {"North", "N-E", "East", "S-E", "South", "S-W", "West", "N-W"},
        name = "8-11 m/s",
        marker = new Marker{
            color = "rgb(158,154,200)"
        }
    };

var areaTrace3 =
    new Area
    {
        r = new [] {40.0, 30.0, 30.0, 35.0, 7.5, 7.5, 32.5, 40.0},
        t = new [] {"North", "N-E", "East", "S-E", "South", "S-W", "West", "N-W"},
        name = "5-8 m/s",
        marker = new Marker{
            color = "rgb(203,201,226)"
        }
    };

var areaTrace4 =
    new Area
    {
        r = new [] {20.0, 7.5, 15.0, 22.5, 2.5, 2.5, 12.5, 22.5},
        t = new [] {"North", "N-E", "East", "S-E", "South", "S-W", "West", "N-W"},
        name = "< 5 m/s",
        marker = new  Marker{
            color = "rgb(242,240,247)"
        }
    };

var areaLayout = new Layout.Layout
{
    title = "Wind Speed Distribution in Laurel, NE",
    font = new Font
    {
        size = 16
    },
    legend = new Legend
    {
        font = new Font
        {
            size = 16
        }
    },
    radialaxis = new Radialaxis
    {
        ticksuffix = "%"

    },
    orientation = 270
};

chart = Chart.Plot(new [] {areaTrace1, areaTrace2, areaTrace3, areaTrace4});
chart.WithLayout(areaLayout);
display(chart);