从Flask API将参数传递给Bokeh autoload_server [英] Passing arguments to Bokeh autoload_server from Flask api

问题描述

我上周开始使用Bokeh，这是一个非常新的东西，我正尝试使用Flask API中嵌入的滑块和下拉列表创建交互式条形图，因此我为其创建了flask api，其显示了带有滑块和下拉菜单，但在更改滑块/下拉菜单值时不会动态更新图表.

I started using Bokeh last week, so very much new to it, I am trying to create an interactive bar chart using sliders and dropdowns embedded in Flask API, so I created the flask api for the same, its showing the chart with a slider and dropdown but its not dynamically updating the chart upon changing the slider/dropdown value.

然后进一步研究发现，我需要为交互部分运行一个单独的bokeh服务器，并从我的Flask api调用自动加载服务器.但是由于我的输入数据来自带有参数作为用户输入的外部API，因此我不确定如何将http post参数发送到bokeh服务器.

Then upon further research I found out that I need to have a separate bokeh server running for the interaction part and call the autoload server from my Flask api. But in that I am not sure how to send my http post params to the bokeh server as my input data is from an external API with params as user input.

script=autoload_server(model=None,app_path="/bokeh-sliders",url="http://localhost:5006")
return render_template('hello.html',script=script)

参考将URL参数从Flask发送到Bokeh服务器，因为我无法对此进行评论，因此似乎集成了该功能以将参数传递给自动加载服务器，但我似乎找不到任何文档.请帮我解决这个问题.

In reference to Sending URL parameter from Flask to a Bokeh server as I am not able to comment in it, it seems the feature has been integrated to pass arguments to the autoload server but I cant seem to find any documentation on it. Please help me in figuring this out.

请注意，请注意，如果不运行bokeh服务器，就不能仅在烧瓶api中进行诸如滑块，下拉菜单之类的交互操作.

On a side note, just to be sure, is it not possible to do interactions such as slider, dropdown, etc in just the flask api without running a bokeh server.

谢谢.

推荐答案

我遇到了同样的问题，无法添加与Flask的交互，并且走了同样的路. 此处还讨论了传递参数的问题.

I was having the same problems, not being able to add interactions with Flask, and went down the same road. The issue of passing arguments is also discussed here.

此功能已添加到Bokeh 0.12.7中，现在您可以使用arguments参数将键/值的字典传递到应用脚本中:

The functionality has been added to Bokeh 0.12.7, and you can now pass a dictionary of key/values to include to the app script using the arguments parameter:

script = server_document("https://example.com/myapp",
                         arguments={'foo': 'bar'})

请注意，server_document是最近添加的，更简单的替代autoload_server

Note that server_document is a recently-added, simpler replacement for autoload_server

对于0.12.7之前的版本，您还可以使用以下解决方法(贷记到github上的kevinsa5):

For versions before 0.12.7, you can also use the following workaround (credit goes to kevinsa5 on github):

@app.route('/amped')
def amped():
    script = autoload_server(model = None, app_path="/amped")
    # `script` is a string that looks like this (the first character is a newline):
    """
<script
    src="http://localhost:5006/amped/autoload.js?bokeh-autoload-element=6b813263-05df-45a5-bd91-e25c5e53c020"
    id="6b813263-05df-45a5-bd91-e25c5e53c020"
    data-bokeh-model-id=""
    data-bokeh-doc-id=""
></script>
"""
    # so to add on the necessary parameters, we have to insert them manually.  hopefully we won't need to urlencode anything.
    # note that request.args = a MultiDict, so be careful of duplicate params
    # http://werkzeug.pocoo.org/docs/0.11/datastructures/#werkzeug.datastructures.MultiDict

    script_list = script.split("\n")
    script_list[2] = script_list[2][:-1]
    for key in request.args:
        script_list[2] = script_list[2] + "&{}={}".format(key, request.args[key])
    script_list[2] = script_list[2] + '"'
    script = "\n".join(script_list)
    return render_template("amped.html", script = script)

这使您可以使用

doc.session_context.request.arguments

这篇关于从Flask API将参数传递给Bokeh autoload_server的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！