Where is the best place to calculate correlation on a set of 18 thousand records in Excel?

Question

Database

There are data of survey clients for 7 years, about 20 thousand records. They are all grouped in a spreadsheet in Excel. If you remove the unnecessary, 4 columns come out: ID, date, features and result. Some people have gone through more than once (i.e. IDs are identical, values ​​in other columns may be different). We need to calculate the correlation. Suppose a person who comes and has certain characteristics, with what probability will he receive a certain result? And of those who came more than once with certain characteristics, what percentage will get such a result and for what time?
I didn’t work much with Excel, I think it has such functions. What is the best way to calculate the data? Should I do this in Excel itself (then I would be very grateful for a link to a tutorial on how to do this) or is it better to convert the database to the same SQL or any of No-SQL and then work further? Then which one to choose, which would be faster in learning and easier to use? There are some restrictions on working with data. They must be located on a specific machine and in no case should they leave it, that is, online services are excluded.

Answer 1

https://support.office.com/ru-ru/article/%D0%97%D0...
better not in excel

Answer 2

The task is not called "probability search", but "correlation search" (between the result and the contents of the "features" cell). Excel has this function: https://researchbasics.education.uconn.edu/using-e...
there would be text notes, but there would be certain result codes and "feature" codes. If, for example, "features" can contain sets of several properties, the task grows and becomes more complicated. In any case, it's best to go to someone who knows how to do such things, because the text of the question implies that you have never performed a statistical analysis,