Now using markdown-interpreted (through ipython) output of tabulate package from compute cells for tables. Removes issues around latex output for dataframes and leads to more coherent output.
Extended due to extended data availability.