Reference one whole table with another - mult "views"

you can use a lookup from the original source table. Can you post an example of the data and what you would ike a summary table to look like?

OK. the answer is yes you can reference another table cell by cell:

to do this make a new, empty table:

In one of the cells (I started at the top left, cell A1) and type the "=", then click the cell of the original table you want to reference.

Now type enter to complete editing the first cell. No select cell A1 again and grab (click and hold the little circle at the bottom right of the selected cell), then drag to the right to fill as needed:

then select the portion of the first row that you filled and fill down as needed

Done:

Hi SZ,

Somehow I get the impression from this thread that you have become fixated on a specific method and forgotten what it is you are trying to accomplish.

In your original post you mentioned "I want to be able to keep several different category-combined and/or differently-sorted tables without having to constantly resort or roll-up categories in the same table."

My guess is that you have decided the 'best' or the 'only' way to do this is to replicated the entire table, then do the categorization/sort on that table. Not quite what you asked for originally, and I don't see it as saving a great deal of time or effort.

Are you happy with the current solution, or would you like to explore others?

Regards,

Barry

Returning to your original request: "I want to be able to keep several different category-combined and/or differently-sorted tables without having to constantly resort or roll-up categories in the same table," here's an example that displays three separate 'categories' (here named 'classes') of transaction taken from a main Data table and placed onto separate tables.

The technique uses an Index table, here named "Test," to mark the location of each occurrence of a category. The index table may be separate, as shown here, of may be included in the Data table as a set of (optionally hidden) columns.

The advantage to including the index on the data table is that it is automatically extended as rows are added to the Data table. The disadvantage is that as part of the Data table, it is vulnerable to accidental revision.

There are two active formulas.

One is entered in A2 of Test, and filled right and down from there to fill that table:

=IF(Data::$A=A$1,ROW(),65537)

Results as shown in Test, column A.

The other is entered into cell A2 of Class A (and with a minor modification into the same cell the other Class tables), and filled right and down to fill that table.

=IFERROR(OFFSET(Data :: $A$1,SMALL(Test::$A,(ROW())-1)-1,COLUMN()+0),"")

SMALL picks off the marked occurrences in order, and sets each number -1 as the row offset for OFFSET. The column offset is set by COLUMN(), which returns the number of the column containing the formula.

When SMALL chooses 65537, one more than the largest number of rows in a Numbers table, as the 'next smallest' number, OFFSET generates an error, caught by IFERROR, which returns a null string (as has happened in row 5 of Class A), which appears as a 'blank' cell.

The Class x tables require at least as many rows as there are items in Data that fit that Class.

Regards,

Barry