Adds hiding of columns, replacing missing values, grand summaries, nano plots and group summaries features

[Patowhiz]

@rdstern this is ready for your review.
This PR adds the following in relation to gt R package:

1. Hiding of columns.
2. Nano plots.
3. Replacing missing values.
4. Group summaries.
5. Grand summaries.

It doesn't exhaust the functionalities of the above gt features but attempts to simplify how to use them.

[rdstern]

@Patowhiz the first thing I wanted to check was the hiding of the columns. It is excellent that the sub-dialog is now available from many dialogs. But, in the hiding it is showing the original data file, (which is irrelevant), and not the summary file, which has the column I need to hide.

Looking at more now.

[rdstern]

Looking at more now.

I next tried to replace missing and (of course) got an error - which is the same problem I presume.

[rdstern]

@Patowhiz it is now 3 days since the problem I found, described above. I would have a much better Christmas if you could fix this soon, or at least reply and convince me it is easy to fix. Currently I have the fear that you might return again to only having gt for a presentation data frame, omitting the main use, which is a presentation (summary) table! That complicated the work for many months and I assumed was resolved by the discussions when you visited England in October.

I still live in hope.

[Patowhiz]

@rdstern Thank you for testing this.

As I previously explained to you and the team in the UK, the tables sub-dialog is designed to work with data already in the data book. It uses our custom controls, which only operate on existing data within the data book.

The summary dialogs create an intermediary data frame (which David refers to as a table) that is then used to produce a gt table. The future plan was to extend or create controls that can interpret the R code generated by the main dialog to allow formatting of a data frame that has not yet been produced.

This is why you are correctly encountering the error column does not exist - the intermediary data frame doesn’t include that column. @ChrisMarsh82 is also aware of this limitation. This PR does not aim to create controls that interpret R code to determine the intermediary data frame produced by the main dialogs. Instead, it continues to assume that you are working with columns from the currently selected data frame.

You can test the gt features in this PR using dialogs that work with an existing data frame, such as the presentation or view data frame dialogs.

However, I do acknowledge that sub-dialog options not requiring modifications to the intermediary data frame, such as the themes feature, should work as expected.

Thanks again!

[rdstern]

@Patowhiz David has come to Reading since I wrote the entry below and he confirms your point of view, i.e. that linking your sub-dialog fully to summaries requires more work and that needs more R skills and concepts than can be expected from you. I hope it can be given high priority, perhaps mainly by @volloholic and @lilyclements amnd perhaps Abib, early next year.

@Patowhiz that's very disappointing. We had the team working on the one-variable summarise, two variables summarise and the main Tables > Summaries dialogs since your visit in October. And you, and @ChrisMarsh82 knew all along that the main parts of your sub-dialog would not work for any of those dialogs. I was really hoping that by mid-December our tables nightmare would be over. Is the main nightmare just beginning? Was @volloholic aware of this, following the October discussions?

You seem to be saying that there is quite complicated and important steps that still remain to be done, and that they are outside what you consider to have been your terms of reference? Should I have known this? How do you plan for your sub-dialog to work with the summaries? Do you plan to be involved, at least to supervise, or to delegate? If delegating, then to whom, and when?