在 jupyter notebook 中使用散景小部件更新散景图

updating bokeh plot with a bokeh widget in jupyter notebook

我想使用 jupyter notebook 中的散景小部件来更新散景图。我的(有点古怪)代码如下所示:

from bokeh.plotting import figure
from bokeh.io import output_notebook, push_notebook, show
from bokeh.models import CustomJS, Slider

output_notebook()

power = 0.5
x = [1,2,3]
y = [i**power for i in x]

fig = figure()
plt = fig.circle(x, y)

def update_plot(power):
    x = plt.data_source.data['x']
    plt.data_source.data['y'] = [i**power for i in x]
    push_notebook(handle=bokeh_handle)  


bokeh_handle = show(fig, notebook_handle=True)

##### new notebook cell #####

callback = CustomJS(code="""
if (IPython.notebook.kernel !== undefined) {
    var kernel = IPython.notebook.kernel;
    cmd = "update_plot(" + cb_obj.value + ")";
    kernel.execute(cmd, {}, {});
}
""")

slider = Slider(start=0.1, 
                end=1,
                value=1,
                step=.05,
                title="power",
                callback=callback)
show(slider)

思路是滑块的JS回调调用python函数update_plot(),改变散景图的数据,然后触发push_notebook().

但是,当我移动滑块时,绘图并没有更新,但是 some weird glyphs appear in the upper left corner (see red arrow)。

执行 print(plt.data_source.data['y']) 告诉我回调和 update_plot() 实际上是在滑块移动时调用的。为什么情节没有正确更新?或者我在这里遗漏了什么?

(我知道我可以使用 ipywidgets.interact 做同样的事情,但我想坚持使用散景小部件。)

我通过在 bokeh.layouts.row 布局中显示图形和滑块小部件来按预期更新绘图:

from bokeh.plotting import figure
from bokeh.io import output_notebook, push_notebook, show
from bokeh.models import CustomJS, Slider
from bokeh.layouts import row

output_notebook()

power = 0.5
x = [1,2,3]
y = [i**power for i in x]

fig = figure()
plt = fig.circle(x, y)

def update_plot(power):
    x = plt.data_source.data['x']
    plt.data_source.data['y'] = [i**power for i in x]
    push_notebook(handle=bokeh_handle)  


##### new notebook cell #####

callback = CustomJS(code="""
if (IPython.notebook.kernel !== undefined) {
    var kernel = IPython.notebook.kernel;
    cmd = "update_plot(" + cb_obj.value + ")";
    kernel.execute(cmd, {}, {});
}
""")

slider = Slider(start=0.1, 
                end=1,
                value=1,
                step=.05,
                title="power",
                callback=callback)
bokeh_handle = show(row(fig, slider), notebook_handle=True)