Wrote

Jean,
You have my file.If what you say is true, please post an example of an exponential fit to my data.

Before reconciliation, play Carlos model wrt your data set.
Not exactly Mathcad exponential fit, but will work.
......................
I don't have your data, don't re-post whereas my SS 6179
has been disabled from downloading plugins.
Make a snippet of 10 pairs that you select randomly.
Zillions data or only 10 is same for fit.
Remember: there is no best fit method.
Pearson correlation is pretty conventional.
In fact, Smath Genfit is superior to Mathcad ... explain.
Smath Genfit enables to best fit the low region, sometimes preferred.
You can probably DIY. Search Samples, Genfit, under my name.
I believe it should work in your latest Smath Solver ... 30405.
Take care Reg ... Jean.

Hello

I believe that one of the best fits for the data supplied is in polynomial of degree 9,
with a correlation coefficient of 0.999254.

It seems to me that emissions cannot grow forever and that
they could remain at 2021 levels or even fall.



CLIMATE CHANGE WORKING_CBG_r4.sm (155 KiB) downloaded 26 time(s).


Best Regards

Carlos

Wrote

It seems to me that emissions cannot grow forever and that
they could remain at 2021 levels or even fall.

Indirectly, that's what HadCRUt4 proves from my last updated 2019.
The Global Warming [2010] 0.8 °C dropped to 0.75 °C [2019]
Taking a range of the last data will closely be Verhulst asymptotic.

Jean et. al.,
I would appreciate your just answering the question that was asked about how to use SMATH and not commenting further. If you bothered to read my pdf report that I provided to you, you would see that I discounted all current models for predicting future warming. That was the point of my report. I only provided it because your original response had nothing to do with my question. I see that was a waste of my time. I assume you are an academician from your judgmental responses to every question I ask. I would ask that you not respond to any further questions from me. I am trying to learn SMATH, not receive unrelated and unsolicited judgments.
Dr. J. Regnald Curry

Wrote

I assume you are an academician from your judgmental responses to every question I ask.

Nah, he is just an idiot who learned to calculate in the past.
He posts mathematically and physically incorrect samples.
Most of his engineering calculations are wrong, for several years.
He doesn't use units, against units, has an open war who uses them.
He doesn't care if his replies are related with original question.
He has his own agenda, just to promote himself.
You, your problem or question is not his concern.
He is an old fart who kicked from mathcad forums.
Now unfortunately he is trolling smath forums.
Don't bother to answer him or his replies.
I don't download his samples.
I even don't look or read any of his replies.
They are useless, or false worst of it.
Most of us just ignore him. I mostly insult him.
Don't encourage him to reply for your questions.
His replies will be completely bullshit or not related.

Regards

JReg…

This isn't an SMath solution, but I suggest you look at the software program Curve Expert Professional (CEP), which has more than 90 different regression models built in. See https://www.curveexpert.net/products/curveexpert-professional/

Using your data, you can have CEP tell you which regression curves have the best fit or you can manually decide which regression models to examine. Then, take the one(s) that you like best and write the equation(s) in SMath for further use. I have done this, for example, for reverse engineering nomographs that I run across as a civil engineer. It's a bit of kludge and can be prone to error, but sometimes it takes multiple tools to build something.

Fred

Thanks guys.👍

Wrote

Wrote

So, I have come to the conclusion that there are no built-in curve fitting routines in SMATH that will allow me to do an exponential fit like I do in Mathcad 15. ....

Hi Reg. The solution proposed at the post #4 in the file { CLIMATE CHANGE.sm (19 KiB) downloaded 58 time(s). } only uses an standard SMath plugin, for al_nleqsolve .

Wrote

I assume that means I must write my own since nothing offered so far gives a good fit like the Mathcad exponential fit. ...

In this graph, the solutions of SMath, Maxima and Mathcad are compared. SMath's solution seems as valid to me as Mathcad's.



Best regards.
Alvaro.

Thanks Alvaro,
I wasn’t smart enough to realize you gave me the answer here. The additional information you sent me via PM was really helpful and made me realize you also gave me the answer here.

Thanks again.
Reg