Submission to CRAN

[ramnathv]

This note outlines the key things that need to be done before I can prepare rCharts for submission to CRAN

JSON Dependency

I need to make a call between rjson, jsonlite and RJSONIO. jsonlite is the most feature complete, whereas rjson is lightning fast. I need to write tests to make sure that I can detect regressions, as I change backends.

Package Structure

I need to decide if I want to go ahead with rCharts and rChartsLibraries. This will give me more flexibility and keep the dependencies separate. This decision will have to be made in accordance with what I decide on core libraries.

Core Libraries

I want to keep rCharts lean. The idea will be to support a set of core libraries, which currently would be polycharts, nvd3, rickshaw, uvcharts, morris, dimple, highcharts, echarts and xcharts. Mapping libraries like leaflet and datamaps will be spun off into rChartsMaps or rMaps, while datatables and other table related js libraries will constitute rChartsTables or rTables. Specialized libraries can be bundled into an rChartsCustom package.

Dependencies

I have an RCurl dependency in rCharts. This is primarily required for publishing as gists. I believe I should make this a suggested package and remove it from imports. I also need to switch from plyr to dplyr, both for performance and future compatibility.

Documentation

This is the big beast that is going to be the hardest to tackle. I can easily document the *Plot functions since they are conventional functions. The challenge will be to document the reference class methods, which are hard to document within the current documentation structure.

Vignettes

These are easy. I can create interactive versions of the vignette by using the slidify playground.

Currently, this is a laundry list. I need to prioritize these as soon as I can, add some milestones, delegate work. The target will be to get rCharts on CRAN by end of March.

[patilv]

If there are small elements I can help with, please let me know.

[lcolladotor]

Hi @ramnathv, how are things progressing?

I'm curious because I would like to submit regionReport to Bioconductor (see vignette) which uses interactive tables powered by dTable().