Add stringsAsFactors=FALSE logic when outputting data frames

[llnelson]

Cascades to viz and metric generation (cleanNestedLists, etc)

[llnelson]

20180924 Attempted; side-effects not acceptable, so:

• df$date remains as.character to accommodate dly, mly, yly options; could detect reduce codes and force with as.numeric() when yly
• df$*countMissing remains as.character because lazy coding used for df population re-uses 'flag' name and value variables

Therefore:

• consuming code needs to convert df$date and df$countMissing if used in graphs

[llnelson]

Review Understanding factors