> L= 1006.536
> L= 1006.537
> L= 1006.535
> It appears to have chosen step size 0.001, not 0.1. It should be
> getting adequate accuracy in both 1st and 2nd derivatives.
> Those little ripples you see in the plot are not relevant.
I'm impressed.
But you're still wrong.
Try this:
-
#n
On 12/03/2020 7:25 p.m., Abby Spurdle wrote:
There is nothing in that plot to indicate that the result given by
optim() should be accepted as optimal. The numerical approximation to
the derivative is 0.055851 everywhere in your graph
That wasn't how I intended the plot to be interpreted.
By de
> There is nothing in that plot to indicate that the result given by
> optim() should be accepted as optimal. The numerical approximation to
> the derivative is 0.055851 everywhere in your graph
That wasn't how I intended the plot to be interpreted.
By default, the step size (in x) is 1e-5, which
On 12/03/2020 1:22 p.m., Abby Spurdle wrote:
I'm sorry, Duncan.
But I disagree.
This is not a "bug" in optim function, as such.
(Or at least, there's nothing in this discussion to suggest that there's a bug).
But rather a floating point arithmetic related problem.
The OP's function looks simple
Hi Abby: Either way, thanks for your efforts with the derivative plot.
Note that John Nash is a SERIOUS EXPERT in optimization so I would just go
by what he
said earlier. Also, I don't want to speak for Duncan but I have a feeling
that he meant "inadequacy" in the CG
method rather than a bug in
Hi,
Why do you want to re-scale RMSE to 0-1? You can change ylim=(0,1) to
ylim=(0, 4600). You may use VEcv (Variance explained by predictive models
based on cross-validation) that ranges from 0 to 100% instead. It can be
calculated using vecv function in library(spm) or you can convert RMSE to
VEc
Thank you all - that was helpful. I guess I assumes everything that got pushed
up to the server was already compiled.
-Original Message-
From: Duncan Murdoch
Sent: Wednesday, March 11, 2020 10:10 AM
To: reichm...@sbcglobal.net; r-help@r-project.org
Subject: Re: [R] Installing Packages
Thanks Hasan and Rui
Rui, as you mentioned
As for the second question, if your RMSE vector had values in the range
2900 to 4600 and the y axis limits are c(0, 1), how can you expect to
see anything?
Then what should be the values of ylim in boxplots? I need to show them as
boxplot between 0-1 or
Hello,
To rescale data so that their values are between 0 and 1, use this function:
scale01 <- function(x, na.rm = FALSE){
(x - min(x, na.rm = na.rm))/(max(x, na.rm = na.rm) - min(x, na.rm =
na.rm))
}
x <- c(SVM=3500,
ANN=4600,
R.Forest=2900)
scale01(x)
# SVM ANN
> (1) An exact solution can be derived quickly
Please disregard note (1) above.
I'm not sure if it was right.
And one more comment:
The conjugate gradient method is an established method.
So the question is, is the optim function applying this method or not...
And assuming that it is, then R is
Hi
I have a regression based data where I get the RMSE results as:
SVM=3500
ANN=4600
R.Forest=2900
I want to know how can I make it so that its values comes as 0-1
I plot the boxplot for it to indicate their RMSE values and used,
ylim=(0,1), but the boxplot which works for RMSE values like 3500
I'm sorry, Duncan.
But I disagree.
This is not a "bug" in optim function, as such.
(Or at least, there's nothing in this discussion to suggest that there's a bug).
But rather a floating point arithmetic related problem.
The OP's function looks simple enough, at first glance.
But it's not.
Plotti
Thank you!
On Thu, Mar 12, 2020 at 12:47 PM Abby Spurdle wrote:
>
> The plotting character is determined by the pch argument.
> i.e. You need to specify the pch argument in the legend call.
>
> Here's the R code to preview the first 20 plotting characters:
>
> > plot (1:20, rep (0, 20), pch=1:20,
The plotting character is determined by the pch argument.
i.e. You need to specify the pch argument in the legend call.
Here's the R code to preview the first 20 plotting characters:
> plot (1:20, rep (0, 20), pch=1:20, ylim = c (-5, 5) )
> text (1:20, rep (1, 20), 1:20)
Empty circles, have a pc
Hi Ana,
I've already given you an example of using the text function with vectors.
I note that this thread contains a lot of duplication.
I'd recommend people read the whole thread before posting.
On Fri, Mar 13, 2020 at 3:42 AM Ana Marija wrote:
>
> Hi Michael,
>
> can you please send me a li
> Alexey Shipunov
> on Tue, 18 Feb 2020 14:34:48 +0900 writes:
> Thank you for the detailed explanation. I tend to agree. However, this
> behavior is relatively easy to remediate:
> This is the piece of the current code:
> ===
> if (!(is.null(labels) && is.null(g
I could make legend via this:
qq(fdr2_sorted$FDR.q.val2, main = "RG_All", pch = 16,
col=fdr1_sorted$group, cex = 0.8, las = 1)
legend('topleft', legend = c('up-regulated', 'down-regulated'), fill =
c('red', 'blue'),bty="o")
but this gives me squares in legend. How do I write this code in order
to
Also how would I add legend to this plot?
I searched qqman pages and there is no mention of that
qq(fdr2_sorted$FDR.q.val2, main = "RG_All", pch = 16,
col=fdr1_sorted$group, cex = 0.8, las = 1)
On Thu, Mar 12, 2020 at 9:30 AM Ana Marija wrote:
>
> Hi Michael,
>
> can you please send me a line o
Thanks for the replies. Since I was seeing this glitch with CG in my 1d and
2d formulation of the problem I was trying to figure out what was going on
that led to the failure. I'll switch to a more suitable method and keep
these considerations in mind.
On Thu, Mar 12, 2020, 9:23 AM J C Nash wrot
Hi Michael,
can you please send me a line of code showing how it would be done.
Thanks
Ana
On Thu, Mar 12, 2020 at 9:16 AM Michael Dewey wrote:
>
> Dear Ana
>
> You can specify the first three parameters to text() as vectors so it is
> all done in one call. That may or may not answer your quest
It looks like a bug in the CG method. The other methods in optim() all
work fine. CG is documented to be a good choice in high dimensions; why
did you choose it for a 1 dim problem?
Duncan Murdoch
On 12/03/2020 2:30 a.m., Skyler Saleebyan wrote:
I am trying to familiarize myself with optim(
As author of CG (at least the code that was used to build it), I can say I was
never happy with that code. Rcgmin is the replacement I wrote, and I believe
that
could still be improved.
BUT:
- you have a 1D optimization. Use Brent method and supply bounds.
- I never intended CG (or BFGS or Ne
It is possible to work out this problem explicitly. Playing with a few
different calls to optim shows that the method="L-BFGS-B" gives the correct
answer.
I don't have particular insight into why method="CG" is problematic.
On Thu, Mar 12, 2020 at 4:12 PM Jeff Newmiller
wrote:
> The help file p
Dear Ana
You can specify the first three parameters to text() as vectors so it is
all done in one call. That may or may not answer your question.
Michael
On 12/03/2020 14:08, Ana Marija wrote:
HI David,
thank you for getting back to me.
Is there is a way for qq() to pick up text label name
The help file points out that CG is "fragile" ... and I would expect that
failing to define a gradient function will exacerbate that.
I think you should use a different algorithm or specify a gradient function.
You might also consider working with the more recent optimr package contributed
by D
HI David,
thank you for getting back to me.
Is there is a way for qq() to pick up text label names on its own or I
have to specify each one manually?
like in this example:
text( 2, 6, "arbitrary")
this is dput for
>a=head(fdr2_sorted)
> dput(a)
structure(list(NAME = c("GO_DNA_PACKAGING_COMPLEX"
I am trying to familiarize myself with optim() with a relatively simple
maximization.
Description:
L and K are two terms which are constrained to add up to a total 10
(with respective weights to each). To map this constraint I plugged K into
the function (to make this as simple as possible.)
Full schedule is available on developer.r-project.org.
Notice that Copenhagen Business School, like all Danish educational
institutions, is physically locked down for two weeks due to COVID-19. The
schedule is automated, but there may be irregularities with the nightly builds,
if things act up
28 matches
Mail list logo