CODEWITHSUNDEEP
chartsbokeh
shaz
shaz

Reputation: 31

Multiple charts on Bokeh figure

I want to add markers to my Bokeh chart, here's my sample code, signals table has positions column which has values 1 and 0, if positions==1 I need to add marker to my TimeSeries chart. How do I do this without using legacy matplotlib plotting, but just using Bokeh interface?

def bokeh_chart(symbol, bars, signals, returns):

    xyvalues = pd.DataFrame({
        "Price": bars['Close'],
        "Date": bars.index.values,
        "short_mavg": signals['short_mavg'],
        "long_mavg": signals['long_mavg']})


    pt = TimeSeries(xyvalues, index='Date', legend=True,
        title=symbol , ylabel='Stock Prices', width=400, height=200)

    #for Scatter, is it possible to be have a dataframe for x and y parameter (1st and 2nd parameters in the function below)?
    p = Scatter(signals, signals.ix[signals.positions == 1.0].index, signals.short_mavg[signals.positions == 1.0], marker='triangle')

    script,div=components(p)
    return {"script":script, "div":div}

Upvotes: 3

Views: 891

Answers (1)

queise
queise

Reputation: 2416

To combine the charts timeseries with a scatter, a solution that worked for me (using Bokeh 0.11.1) is to use basic glyphs instead of charts. In your example you would use the argument of figure x_axis_type="datetime" and the basic glyph line instead of the chart TimeSeries, and use the glyph triangle instead of the chart Scatter. 

from bokeh.plotting import figure

p = figure(title="My title", width=400, height=200,
           x_axis_type="datetime")

p.yaxis.axis_label = "Stock Prices"
p.yaxis.axis_label = "Date"

p.triangle(xyvalues['Date'], xyvalues['short_mavg'], legend="short")

p.line(xyvalues['Date'], xyvalues['Price'], legend="price")

I hope it also helps in your case.

Upvotes: 1

Related Questions