Comments and Ratings

Less than 2nd rate. Why bother with this, if VP can't be bothered to make it easy to use? Not a uicontrol anyway. Just as easy to use use ginput if you must create a new figure window.

John wrote:
///First, I see that when the cursor drops outside the axes, you get a NaN when you go below the lower bound. But there is no constraint on the upper end. This seems inconsistent. Why do it this way? ///

SLIDDER is a part of a GUI which integrates several matrix-oriented procedures, all need NONNEGATIVE integer vertical and/or horizontal parameters. NaN means that corresponding procedure skips the processing in corresponding matrix dimension. This is a very convenient and clearly percepted convention.

Therefore I do not need to treat the upper bound in a similar way. Moreover, the filtering windows vary in the range from 10 to 100. The greater the upper bound, the more difficult is to catch wanted window. Therefore the option of overriding the upper bound really makes the users life easier. The thing is that the SLIDDER works in parallel to an editable text box which receives the output from the SLIDDER. If catching numbers with SLIDDER appears difficult, the user has to type the windows by hand. However, the probability of a conflict between x-y sequence convention and 1-st and 2-nd dimension convention increases, because not all the users eat wooden chips. This probability is minimized by the option of overriding upper bound: if my default is 31 31, I still can easily reach much greater values like 71 or so with the SLIDDER. Therefore accepting upper bound smaller than it finally appears to be necessary improves catching numbers and leads to no conflicts exactly due to overriding the upper bound.

Thus, what seems to be inconsistent in the West, can easily appear to be a life necessity in the East.

Because not always the parameters should be integer I have added the float mode. In my practice, I did never work with fixed steps < 1. However, if such a special case comes, then the continuous output can easily be quantified. However, should I add this option, any expert, not to speak about normal users, will be overloaded. Too good is again bad.
  

///The simple fix is to define the "CloseRequestFcn" as the same as clicking on the finish button in the window.///

This is a really useful and helpful remark, thanks. I shall try and probably resubmit the SLIDDER if not delete it. I personally do not lose anything.

 The name SWIDDER does not mean SaintWIDDER. Another John informed me that my submission can not be accepted because it was compressed by .rar, not .zip. I had to resubmit, but it was too late and I was tired. In a couple of days many people come here to participate in the conference, and those who have seen the SLIDDER before coming (as well as MLIN, POWEL, LINPAT which are parts of the same GUI) will have more preliminary informations, therefore they will use their time here more efficiently.

I'm closer to Shaun here than to Duane. I could potentially see someone using this as a graphical input utility.

This code has its subordinate functions included. It has help, that makes some sense. I was able to read the help and figure out how to use the code with no difficulties.

I did find a few minor issues.

First, I see that when the cursor drops outside the axes, you get a NaN when you go below the lower bound. But there is no constraint on the upper end. This seems inconsistent. Why do it this way?

Next, in order to create an slidder that lives on an integer domain, you just use integers as endpoints. A floating point one happens when the endpoints are not integers. But suppose I wanted one that goes in steps of 0.5 on the y axis, and 0.25 on the x axis? Yes, I could shift/scale/transform the output, but that would be a kludge, since the axis labels would be all wrong. Better would be an optional step argument for the respective axes.

I also got the code to give an error when I closed the window instead of clicking on the finish button. The simple fix is to define the "CloseRequestFcn" as the same as clicking on the finish button in the window.

My final comment is a cosmetic one. I see that the web site description calls this "swidder" in several places, yet the code is named slidder. A mid-stream name change perhaps?

As for my rating, I was tempted to call this a 3 (fair). But I do think that Duane has been too severe in his review, so I considered giving it a 4 rating. I'll not assign a numerical rating at all at this moment, planning on returning to it later.

Another amazing program from V.P. I never understand your programs, but they make me smile. 5 stars for making me smile.

You have astonished me and partly convinced. Therefore a special offer to you: just delete function CHEER. LINPAT will still work.

Let the users decide whether the difference between the following principally equivalent expressions is that important.

r=unifrnd(A,B,M,N);

r=A+(B-A)*rand(M,N);

Of course, if B<A, then UNIFRND fails. Do you mean that this is a sufficient reason to avoid using UNIFRND?

In any case those who have no STATS tlbx, can easily replace unifrnd if necessary. What is the percentage of such users?

very good

You can make a hexagon and many other shapes using my STAR. It can also plot the line or patch around point(s) which you specify.

It is not difficult to include z0 into list of inputs. However this can be easily done by those who really need it. I am too lazy to supply trivial options.

You can make a hexagon and many other shapes using my STAR. It can also plot the line or patch around point(s) which you specify.

It is not difficult to include z0 into list of inputs. However this can be easily done by those who really need it. I am too lazy to supply trivial options.

I see in the last days something has happened. Some unknown people claim to be experts without having compared the quality of the results with any other method of quadratic interpolation. Quadratic spline does the job better, if the points are relatively closely positioned. But I do not claim that this program is better than spline.

The help is impossible to read.

It seems silly to force the user to also download cline, which does not do anything better than plot does.

Omg, this algorithm is wrong!!!!
try for
simpsum([1,0,1,4,9],1)

it return 2.333

Yes, I recognize that in both cases, in CUMSIM and in CUMSPLINE, a simple vectorization can be done. For instance, in cumspline all commands starting from the first FOR loop, can be substituted by 1-liner
which represents the proper Simpson integration block:
CS(2:ROWS,:)=cumsum((Y(1:2:N-2,:)+4*Y(2:2:N-1,:)+Y(3:2:N,:)).*repmat(diff(x(:)),1,COLS)/6);

Somehow intuitively I did not like it, but now I have checked it.
Indeed, if Y is only one column, this saves about 30% time, as John sais.

However, if I have many columns, such an improvement can act according to the proverb "better is an enemy of good". Take an example
 x=sort(5000*rand(1,10000))';
y=[sin(x) cos(x) exp(-x)];
y=repmat(y,1,20);
[CS,CS2]=cumspline(x,y);
t = 0.172 0.156
The first time here is the calc. time of vectorized version, the second one is the time of the criticized version, which appears
to be better than that of the vectorized one!

 I am not sure, whether such a vectorization is meaningful here. Again, we have relatively big data arrays (10000*60), and we discuss the time differences about 0.02 sec. Does it make sense?

Here my timing in CUMSIM, if any further questions:

just after interp1:

tic
CS2(2:ROWS,:)=cumsum((Y(1:2:N-2,:)+4*Y(2:2:N-1,:)+Y(3:2:N,:)).*repmat(diff(x(:)),1,COLS)/6);
t(1)=toc;

tic
for i=1:3
 IND(:,i)=i:2:N+i-3;
end
FIL=[1 4 1]'/6;
  
CS=zeros(ROWS,COLS);
D=diff(x);

for i=1:COLS
    w=Y(:,i);
CS(2:ROWS,i)=(w(IND)*FIL).*D;
end
CS=cumsum(CS);

t(2)=toc

.

I have seen only CUMSIMPSON in "integration" as a comparable alternative to CUMSIM.

 Another program, CUMSIMPSUM works only at equally spaced data and cannot be directly compared in the case of irregularly spaced data. CUMSPLINE was written by me to show that the results depend on quality of interpolation, submitted together with CUMSIM. I pretend however that CUMSIM is a TOP QUALITY product within the frames of QUADRATIC INTERPOLATION. What is the third alternative? If you vote for deletion of CUMSIM, I expect that you know something better, other possible reasons would be not objectively correct.

Hey John!
I have submitted this file on monday, before the PARFIL has appeared on FEX!!!

Now what does your voting for deletion mean? You don't like that my program works better than that of your friend? You are a battler, not me. Sorry, but I do not like to be accused for the crimes done by others.

.

Hi Duane.

I can answer all your questions, but I need some time for that, not right now. I do not understand, what are your difficulties.

If you want to find the interpolated y-value in the middle of given internal interval, you can use two given x,y points, defining the interval. And you need one more point for quadratic interpolation. You can select it either before, or after the interval. Thus you have two different possibilities, and there are no general limitations which could oblige you to select one of them and ignore the second one. I have expressed this shortly as "equal in rights", which I expected to be sufficiently clear.

What I do not understand, how can one decide whether something is of some value or not without any attempt to understand, what is it. Such an aggressivity is unusual for civilized people, and it is not helpful for any real discussions.

I think, enough at the moment.

Hi John!

The first real expert sheding some useful light. Thanks for your citation. In fact it supports my position here. Moreover, we do not need to go too far outside ML community, and in many places ML doc uses the definition: order = number of all possible polynomial coefficients at given degree, and polynomial degree = highest degree of x, so that ORDER=DEGREE+1. Because of this strong binding I am usually relactant, which of the two could be better used. In any case, if one uses some definitions, not accepted in ML, he should have specified this at once, which is not the case. I wonder why the discussion is around some things of 1000-th importance instead of essentials, but if a question is put, I feel obliged to answer. I am finishing now based on PARFIL general quadratic interpolation (vectors or matrices) like

interp1(x,y,X,'quadratic')

with arbitrary X. Since I have published INTERPOL, which has shown a gap in interpolation, i.e. missing classical quadratic and cubic interpolations, I did hope that this gap will be closed, but to my knowledge this has still not happened.

I don't know who is Jos. A serious remark would be: so and so % calculation time can be saved with long data by avoiding FOR or FIND. Other arguments are highly incorrect, reflecting some strange emotions. I have enough emotional "experts" here.

>>It is clear to me that we have a language issue here.

A small correction: YOU have a problem, not me. And here is the proof:

>>Our understanding of polynomials differs and I now know what you mean by "order" and "degree".

Exactly this is "THE LANGUAGE PROBLEM". I use everywhere accepted definition, you have your own, apparently, a different one, as you have demonstrated. I am not going to teach you English or math. In turn, I have strong doubts that I can learn anything from you.

>> Is added x data point (x(k)+x(k+1))/2?
Yes. This is clearly written above. Why again asking?

 >>Is parabolic equal to quadratic, meaning a second order polynomial?

Where did you see QUADRATIC PARABOLAS as SECOND ORDER polynomials? QUADRATIC polynomials have order three and degree two. In cubic polynomials order =4, degree=3.

In internal intervals there are TWO EQUAL IN RIGHTS possibilities for QUADRATIC INTERPOLATION. I am used to respect rights. This explaines high accuracy of INTERPOL, although it works only with equally spaced data. In PARFIL
I do ONLY ONCE BOTH POSSIBLE QUADRATIC INTERPOLATIONS in every internal interval, subsequently only using these results.

In first and last intervals the story is somewhat more complicated, and, as IS WRITTEN IN HELP, I have included a trivial option for cubic interpolation ONLY IN THESE INTERVALS. By default, QUADRATIC interpolation is made even here.

Enjoy my results. I shall accept your voting for deletion, if you succeed a better QUADRATIC interpolation at the same calculational price. It is known, interpolation of any degree can be reached as repetitive linear interpolation of previous results. You don't need to suspect such cheap tricks, but it seems so.

... and remove all of the required toolboxes. They are *not* required, and it is misleading.

Hi Duane!

PARFIL is submitted for those ML users who are interested in good routines. Sure, I shall be happy if there will be many such users.

All TMW products appeared occasionally, in fact none is necessary. Therefore take it easy. Next time I shall be more careful submitting a file. Help is very detailed, perhaps even more than necessary. Just try any reasonable example and judge yourself.