Major_Tom Disproves NIST Claims in a Number of Key Areas

LOL. Exactly how are you going to get a velocity plot and an acceleration plot by running the filter once (without recoding it so that multiple derivative orders are output) ? :rolleyes:
Let N be the degree of the best-fit polynomial, and let W be the number of points in each window.
Code:
for each window:
  for i = 1 to N:
    compute derivative i
What part of not coding specifically to force the behaviour didn't compute ?

What language is that ? Appears at first glance to be psuedo-code.

With modern computers, and time series as short as the ones we've been discussing in this subforum, it's beyond silly for femr2 to worry about the speed of the Savitzky-Golay algorithm.
You're making it up as you go along again. I make no such assertion.

Easily done. Run the Savitzky-Golay algorithm over the displacement data, generating both velocity and acceleration data in that one pass (as described above).
:rolleyes: by recode. Nothing like changing the goalposts to suit your desire to whine eh.

femr2 said:
derivative order cannot exceed polynomial order
Incorrect.

Polynomial functions are C∞ (smooth, infinitely differentiable).
Ye gads man. Talk about pedantic.

Whilst you're correct in mathematical purity, you'll find that most implementations enforce the limit, as you would expect...unless you WANT a whole bag of zero :rolleyes:

Of course taking the second derivative also forces you to use a poly order of at least 2.
Incorrect.
Again, see above :rolleyes:

(The second derivatives of linear and constant-valued functions aren't very interesting, but they do exist.)
"aren't very interesting" ?! LOL. Care to elaborate, or say zero ? :rolleyes:
 
I've highlighted in yellow the part that proves you're strawmanning. A bigger window size could compensate for the unfiltered noise due to using a higher degree polynomial, with the net result of having an equivalent smoothing.
Not strawmanning at all. Highlighting that increasing poly order tends to decrease smoothing. Somewhat skeptical of the true honesty of your subsequent responses. But hey ho.

I didn't suggest window size to remain constant.
I didn't say you did.

Ok, that's an implementation detail.
Indeed, a point WDC seems to feel doesn't apply to him, nor his desire to finally write "incorrect" a few times :)

Your switching of terms is what confused me.
Say wha' ?

You were implying that raising the degree may not be desirable
Correct.

Which one is better: 1 pass or 2 pass?
Personally, I think my 2-pass curve is more useful than the 1-pass curve I posted, especially around the start of descent.

Anal folk may well whine ad infinitum about mathematical purity, however, the bluster still boils down to speculation and opinion.

Perhaps Oystein's "god data" may settle the base arguments about whether these methods can reveal useful acceleration trend information from displacement data containing noise, or the accuracy of such.
 
Last edited:
femr2- Iirc single pass was far too noisy for my purposes,
We can't know whether that's correct until femr2 explains his purposes.
Major_Tom These pissing contests in this forum can go on for years. All one gets as a result is a bunch of urine.
femr2 - Anal folk may well whine ad infinitum about mathematical purity, however, the bluster still boils down to speculation and opinion.
I believe W.D.Clinger may be asking for the purpose in order to first determine an appropriate smoothing model to be selected that will yield this precision goal (it might not be SG, or SG with appropriate parameters), and the accuracy matched to this precision.


........................................

Just to be clear, if some guy presented measurements better than those I am using, I'd say, "Thank you", and include their work when constructing the mappings. I do not care if they are a "debunker" or "truther" or if the guy has little green antennas on his head.

Nobody can "debunk" the mappings as a whole because I would just absorb their information and link to their work. Any "debunking" of specific features of the collapse processes only helps me produce mappings of an even higher quality.

It doesn’t seem MT and femr2 are grateful for W.D.Clinger's expert help after all.
 
I believe W.D.Clinger may be asking for the purpose in order to first determine an appropriate smoothing model to be selected that will yield this precision goal (it might not be SG, or SG with appropriate parameters), and the accuracy matched to this precision.
To steal a phrase, more to the point the intent is clearly "prove femr wrong version the umpteenth".

The "discussion" is still about a graph produced early 2011.

Plenty time to have supersceded the irrelevant femr2 methinks, and provided a wealth of "superior" data using whatever methods "others" think would be more appropriate.

The base "purpose" has been clear for YEARS...reveal the acceleration behaviour of (the NW corner of) WTC7 in as much detail as possible.

It doesn’t seem MT and femr2 are grateful for W.D.Clinger's expert help after all.
He's not "helping".

Will is more than welcome to present his own data and derivatives.

As MT said over on the appropriate thread...
Thanks for those graphs.
...
W.D. Clinger, of those results in your last post, which one do you think most accurately depicts the acceleration trend of the NW corner of WTC7 ?

There's a useful question. :)
 
Last edited:
What degree and window size did you use for 1 pass?
:)

2, 51

I suggest "others" don't make assumptions about what parameters they "think" may have been used on other graphs. Assumption is the mother of all...

(Though it can be entertaining)
 
Not strawmanning at all. Highlighting that increasing poly order tends to decrease smoothing. Somewhat skeptical of the true honesty of your subsequent responses. But hey ho.
And failing to address my concern that you can implicitly be raising the poly order by applying the smoothing twice, by throwing a red herring. Speaking of honesty in responses.


Anal folk may well whine ad infinitum about mathematical purity, however, the bluster still boils down to speculation and opinion.
The base "purpose" has been clear for YEARS...reveal the acceleration behaviour of (the NW corner of) WTC7 in as much detail as possible.
I guess you don't see a conflict between these two statements. I do.


Which one is better: 1 pass or 2 pass? What degree and window size did you use for 1 pass? And for each pass in the 2 pass version?
Personally, I think my 2-pass curve is more useful than the 1-pass curve I posted, especially around the start of descent.
Three questions, two answers, and...
I suggest "others" don't make assumptions about what parameters they "think" may have been used on other graphs. Assumption is the mother of all...

(Though it can be entertaining)
... such a childish reaction.

Maybe if you're asked with "pretty please with sugar on top"? (speaking of pirate references).


Your switching of terms is what confused me.
Say wha' ?
Say that you were switching the meaning of "pass" from the theoretical equivalent to "iteration", which is what has been used in the thread, to your own concept of what it should mean, which is more like "run of the program" in practical terms. Just another instance highlighting tfk's criticism about how you confuse people by your nonstandard use of terms, something that collaborates in introducing confusion to the readers of your posts together with your withholding of data (see above). I see that lack of willingness to ease comprehension to others as rude, for the record.
 
And failing to address my concern that you can implicitly be raising the poly order by applying the smoothing twice
Less smoothing is fine and dandy by me. No smoothing at all would be great, but that's simply impractical.

I guess you don't see a conflict between these two statements.
Correct.

Ho hum.

Three questions, two answers, and...

... such a childish reaction.
You mean "thank-you femr2 for providing me with information requested" :)

Say that you were switching the meaning of "pass"...
No switching going on my end. Perhaps you've decided to jump upon WDC's decision to recode what would normally be within a main "SavitzkyGolayFilter" function, in order to a) make a pedantic "score" and b) output multiple derivatives in a single call. Needless to say most implementations request you provide the single derivative order you require.

Just another instance highlighting tfk's criticism about how you confuse people by your nonstandard use of terms, something that collaborates in introducing confusion to the readers of your posts
Nah. No switching going on my end. Looks like you're simply confusing yourself. But I'm sure you must feel great thinking you've justified a negative comment towards femr2. Awesome. High five :p

I see that lack of willingness to ease comprehension to others as rude, for the record.
I find folk continuing to criticise graphs and data generated years ago utterly hilarious. I find the unending pile of "prove femr wrong version the umpteenth" incredibly rude (especially as the vast and utterly overwhelming majority of such criticism is total and utter testicles) For the record :)
 
Awesome one there, beachnut. That's a keeper.

I take it then that Major Tom is not an engineer? :D

Which is not a trick question or anything. I just think that having this information would greatly influence me in my decision-making whether or not to read this thread from the beginning.
 
Last edited:
You mean "thank-you femr2 for providing me with information requested" :)
"I want to check if A is equivalent to B. Can you give me the parameters for both?"

"For B, 2, 50. For A I won't tell you."

"Thank you femr2 for providing me with information requested."

Nah, doesn't really fit.


No switching going on my end. Looks like you're simply confusing yourself.
Switching from the meaning established along the thread (which is the universally accepted one in the context of the topic at hand) to your own, far less universal definition. Yes you did. Not consciously, probably.


I find the unending pile of "prove femr wrong version the umpteenth" incredibly rude (especially as the vast and utterly overwhelming majority of such criticism is total and utter testicles) For the record :)
Your history of miscommunication and causing confusion dates back at least to the beginning of your postings here. There might be a relationship, mind you.
 

ISF - Join now!

Every member here is approved by hand. No bots, no spam, just people who care about evidence and honest debate.

Membership is free!

Create your free account

Back
Top Bottom