Visual inspection of pandas frames

Question

I work a lot with pandas data files containing timers. For data analysis, I make graphs for a set of columns and for a fragment in the index (datetime).

However, I missed the interactive plot tool or gui, which would allow me:

• click in the list of columns to be built.
• there is a slider for scrolling time
• there is a slider for setting the length of the displayed interval
• have mouse related information with values
• for comparing timeseries with different scaling: either the multiple Y axis in the same plot, or subtitles (connected in the time domain)

Does anyone know a tool that has this functionality? Ideally, this would be integrated into ipython notepad, but a separate tool (using dataframe as input).

Thank!

Answer 1

You can try the charts library . It uses the javascript Highcharts interactive library to quickly create interactive stories. It has some interesting features:

• Interactive diagrams (scaling, slide, hovering over points, ...)
• Used in IPython Notebook
• Directly print pandas dataframes
• Use the interactive variable selector to choose which variables you want to see. (This can be useful so that you can build your framework in one try and quickly view the contents by adding and deleting data).
• Build any data size using the asynchronous build method.
• , .png,.jpg .svg, .
• , (!)

.

, , .

:

Answer 2

Bokeh, , .

, , "" .

:

http://bokeh.pydata.org/en/latest/docs/server_gallery/crossfilter_server.html http://bokeh.pydata.org/en/latest/docs/server_gallery/sliders_server.html

pandas,