I have Reviewed Ferguson's Code - It's a Joke

Armstrong Economics ^ | May, 8, 2020 | Martin Armstrong

[dadfly]

as we from the 70’s used to joke, “this sounds like a big Kludge.”

Gates indeed. some friends of mine used to joke back in the day that they needed to keep “Bill” away from the code.

however, no joking here, unlike Gates and DOS/Windows, this software is killing people.

[Cboldt]

-- The fact that it cannot produce repeatable results for a given set of inputs is the showstopper. --

Seriously. The way I read this, the code is designed to produce different results each time it is run.

Well, okay, that might make sense. The inputs have uncertainty, so the result will necessarily have uncertainty. You'd think the code would then loop and "do the uncertainty math" on the conclusion. THAT result should be repeatable withing a defined uncertainty / Z-value, whatever you choose for your distribution.

At any rate, instead of GIGO, we have garbage in, garbage algorithm, garbage out ... GIGAGO

[Magnum44]

Im dont think we are arguing. Im just trying to stay on point and not get distracted in the minutia like the author does.

This software WRECKED a 20 trillion dollar economy just in this country, not accounting for the rest of the world. No one should be defending it on any level.

[ImJustAnotherOkie]

I never learned the ANY language.

[Blagden Alley]

C++? Those Me-Tarzan, You Jane languages are so primitive.

[FreedomNotSafety]

“ Follow the Science”, they said
“Trump should let the science experts manage the national response”, they said
“Fauci and Birx are the smartest scientists in the US”, they said

Funny, “ they” are still saying this

You are talking about the Fearpers, right?

[silverleaf]

Nope

[semimojo]

Anyone who understands math and modeling should know that you really have NOTHING to support prediction.

Ridiculous. Applying past experience with other diseases to new ones is the essence of epidemiology.

To think we can't learn anything from earlier virus-spread diseases is silly.

You say modeling based on our past experience is useless "NOTHING to support prediction".

Our leaders need to make decisions. How do you propose they inform themselves?

[Campion]

You can only program in ANY on a computer with an ANY key. Everyone knows this. /s

[Campion]

The way I read this, the code is designed to produce different results each time it is run.

To properly regression test, there should be a way to circumvent the randomness so as to get a repeatable answer. Otherwise, how do you distinguish between a breaking change and an expected random result?

[Magnum44]

I am specifically referring to what these political doctors refer to as 'models'. These are math constructs that they use to 'predict' the spread of a disease. These models must be based on data, and almost always in controlled experiments. Real life is not a controlled experiment. So I am not saying you can learn from history. Dont put words in my mouth, please. I am saying you cant 'fit' data, and then use that fit to extrapolate (that is the correct math definition in this case), because your fit model only fits whats behind you. You get tomorrows data and the model parameters change because you added a new data point.

Again, the hockey stick from global warming was the worst example of this at the time. Now I would have to say that this model and its dire predictions have been worse to the world than the hockey stick ever was.

[ImJustAnotherOkie]

That’s kind of QWERKEY.

[epluribus_2]

Wickedly fast and accurate. It still runs on your high end graphic cards. Not for business data processing. Algorithms have been tuned over half a century like modern jet engines

Mostly missing miles of unwicked object oriented abstraction layers.

Also no real recursion in the good parts. Work done in registers not virtual ram. This confuses java thinkers.

[Cboldt]

-- To properly regression test, there should be a way to circumvent the randomness so as to get a repeatable answer. --

Yep. In several places, I gather. I suspect randomness is introduced in more than one place/variable in the chain.

[Dr. Sivana]

Fortran of all things.

Fortran is not the end of the world, particularly for modeling. Heck, we went to the moon on Assembly Language. Good code is good code, and a lot of important stuff has been done in Fortran, and it is still in use all over the place, paricularly in mainframe environments alongside COBOL.

The problem is it was poorly written.

[semimojo]

I am saying you cant 'fit' data, and then use that fit to extrapolate (that is the correct math definition in this case), because your fit model only fits whats behind you...

The epidemiologists use the very limited data they have about the new disease, make some assumptions based on the likelihood it will perform somewhat like similar pathogens, and caveat the hell out of their predictions.

So again, at the beginning of a pandemic with a novel virus how should people make decisions?

[ImJustAnotherOkie]

I would say anyone developing in FORTRAN in these days is a retarded.

I always like the phrase some people get 5 - 10 years experience, but some people get 1 year of experience 5 or 10 times.

[Magnum44]

Its funny. You really want to defend the use of this model by Fauci and others after its been wrong, wrong, wrong, and after the devastation that has been brought down on the American people and the US economy?

What similar pathogens did they base all their other assumptions on?

Feel free to continue your argument with someone else. I have little respect for what you refer to as epidemiologist, or any medical or scientific specialty, who would use 'limited data' to wreck entire societies for their own vanity.

[reasonisfaith]

It’s very clear we don’t have to understand or even look at the details of the code in order to know the whole thing was a sham.

[reasonisfaith]

The answer to your question is FIRST AND FOREMOST they should separate the natural occurrences from the unnatural.

A virus created in a lab is unnatural.