Formula from table with operation restrictions 2 (col2 is only a helper)

[DynV]

This post is basically the same OP as another thread of mine but is of a different subject and has a different 3rd column.
I have a mostly automated profile/sheet/information that I'd like to automate even more as it allow custom formulas using limited data and operations. In the following table the column (col) 1 is entered and col 2 is automated and I'd like to do so for col 3 (reason in previous sentence). In the custom formulas the only information related to col 3 is col 1 & 2 and operations are arithmetic (only +, -, x, /) and rounding: round(), floor(), ceil(). There is no need to figure col 2, it's only there if it helps finding the restricted formula for col 3; although if there's a way to only get it from col 1, it would be a bit more robust.

1	2	4
2	2	4
3	2	4
4	2	4
5	3	6
6	3	6
7	3	6
8	3	6
9	4	6
10	4	6
11	4	8
12	4	8
13	5	8
14	5	8
15	5	8
16	5	8
17	6	10
18	6	10
19	6	10
20	6	10

Step by step: In the profile/sheet/information, a user goes row-by-row, hopefully going all the way to the end but might stop along the way, stopping would have nothing to do with the data. The use currently have to enter col 1 & 3 (as they go along, thus row-by-row), in the profile/sheet/information what's related to col 2 is automated and will always reflect what's in the table (whatever is in col 1, it will become col 2 of the corresponding row). As mentioned currently the data in profile related to col 3 has to be entered manually with the corresponding disadvantages, as it will function incorrectly if a user forget to modify it after changing the data corresponding to col 1, or do so incorrectly. I would like an algorithm to automatically modify the data related to col 3 so there's no chance of a user forgetting to change it after changing col 1; the only operations allowed are in the 4th sentence.

I'm not sure if this is the right forum for this request, moderators please change it if appropriate, or those that know one to send them a suggestion to do so.

Thank you kindly for your help

 

[mmm4444bot]

… I'm not sure if this is the right forum for this request …

Hey there, DynV. You're welcome to post your math-related questions on the boards. The active membership is a varied community, so no worries. These members decide for themselves whether or not to participate in a discussion. Some may take interest in the sequence below, and some of those might post a reply. Others might try sequences involving all your data. Also, there are a few threads in the forum with no response at all.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
4 4 4 4 6 6 6 6 6  6  8  8  8  8  8  8 10 10 10 10

I like solving puzzles. Finding a formula for a sequence can be like a brainteaser, but I don't have time for it until later next week (time off). Maybe I can come up with something. Or, maybe somebody else will find it (or show that it's not possible, given your constraints); hard to say.

Cheers :cool:

 

[mmm4444bot]

So far, I've tried a number of regressions. I had hoped to get lucky using multiple-polynomial regression with all three columns. Each regression gets close, but there's just enough error to cause one pair of adjacent col3 values to round to the wrong Integer. I have some other stuff to try, in a while.

 

[DynV]

A reminder that col2 is only a helper, as in if col1 by itself won't work, and does not have to be included at all (but do so if it help).

 

[DynV]

The solution was shared with me on another math forum.