Goal of storing pandas dataframe on redis server

Community Support
1

First, I’m totally sucked in by the amazing examples of Bokeh that I’ve seen. I sincerely hope development gets some momentum behind it.

I’m trying to learn the structure of the code with a few examples. I am working through IPython exclusively. My question is: What is the intention behind storing the pickled DataFrame object on the server?

Below are two examples, the first one works as expected, the second one does not because using a DataFrame requires connection to the server.

  1. named vectors:
    p = PlotClient()
    p.notebooksources()
    x = np.arange(100) / 6.0
    y = np.sin(x)
    z = np.cos(x)

data_source = p.make_source(idx=range(100), x=x, y=y, z=z)

p.hold(‘off’)
plot1 = p.plot(‘x’, ‘y’, ‘orange’, data_source=data_source)
plot2 = p.plot(‘x’, ‘z’, ‘blue’, data_source=data_source)

grid = p.grid([[plot1,plot2]])
grid

  1. DataFrame
    p = PlotClient()
    p.notebooksources()
    df = pandas.read_csv("…/tests/auto-mpg.csv")
    source = p.make_source(df)
    p.hold(‘off’)
    plot1 = p.scatter(‘mpg’, ‘yr’, data_source=source)
    plot2 = p.scatter(‘mpg’, ‘weight’, data_source=source)
    grid = p.grid([[plot1, plot2]])
    grid

My guess is that the intention is to allow dynamic plots, where the plot updates in real-time as the data is changed. But it would seem to make sense to be able to use a DataFrame in this way to create a static plot. Am I on the right track?

Thanks,
Spencer