Forum Discussion
Is now the time to abandon the concept of relative referencing?
Thanks for the shout out. Those functions look great so I guess it's time to package your LAMBDAs into 5g libraries. 
BTW - 5g is on fire. In December I presented 5g at the University of Arizona's MECC championship. Last month Danielle Stein Fairhurst presented 5g functions a the Excel Global Summit. Model Citizn just posted an article on 5g. All next week I have 5g workshops for Full Stack Modeller. In April I'm at the University of Tennessee. In May I'm presenting at FMI (Financial Modelling Institute). Finomatic Consultancy is converting their models to 5g... and this all traces back to when you introduced me to your work with array formulas, what, 10 years ago? Thanks for all your help Peter.
10 years? It must be about that, the FAST Standard was launched in 2014 during a meeting at the ICAEW. As a heavy user of both defined names and CSE arrays I queried features of the standard but it was not the occasion to follow up the discussion. In the following year I pretty much trained myself to commit all formulas with CSE and I think we were sharing ideas well before EuSpRIG 2015 (not that CSE arrays are that pleasant to work with). I also got invitations to ModelOff 2017 and 2018 Bootcamps and pestered Microsoft staff to improve their array offering in much the same manner as Tables had done for list data.
The complete rework of the Excel calculation engine, when it came, was far more than I had expected and has led us to the point where your pre-built modules are even possible. I may need some guidance from you on testing and publication. The algorithms that interest me most at present are those that use array shaping, sorting and filtering as key elements of the solution. Without arrays as the fundamental building block one could be left floundering in a sea of helper cells.
If you are giving any presentations that I could listen in on, could you send me a link/invite?
- We should zoom.
