create intake source for dataframe plot

[ryanSoley]

Is your enhancement request related to a problem? Please describe

we would like to be able to share visualizations alongside the projects and experiments in the intake catalogs. for example, catalog.read() returns projects and experiments. we'd like to ultimately have something like catalog.serve() that mimics the viz module's serve functionality to launch a dashboard with the experiments returned by catalog.read()

in #441 we created a base source for visualizations. in #442 we created a source for the experiments table. we will need a concrete implementation for each visualization like the one implemented for #442. this issue will track the creation of a source for the dataframe plot

Describe the solution you'd like

add a new DataframePlotDataSource that extends VizDataSourceMixin. the __init__ function should take in any additional arguments specific to the DataframePlot. _get_schema should create a visualization and set it as the _visualization attribute, as well as update the metadata with any required attributes

Additional context

at this point, the visualization stored in _visualization does not need to have any experiments in it - we'll get to that later

this should be implemented very similarly to the solution to #442. it may even be possible to move some common functionality between the ExperimentsTableDataSource and the new DataframePlotDataSource to the base VizDataSourceMixin if its shared across both viz's implementations

[ryanSoley]

closed w/ #465