Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

3 Pages<123
Options
Go to last post Go to first unread
Offline hanskl  
#41 Posted : 29 October 2020 10:36:28(UTC)
hanskl


Rank: Advanced Member

Groups: Registered
Joined: 25/10/2017(UTC)
Posts: 94
Norway

Was thanked: 6 time(s) in 4 post(s)
Originally Posted by: Jean Giraud Go to Quoted Post

Your latest "complete" does not calculate same values ?
Much shorter in real values. For the remaining of the project, all red SS 6179.
Can't even copy/paste algo in same document, neither export.

AlgoDamaged.PNG



The difference is in the input, as well as the modification I have made to c.

Input in my equation is ρ [deg], while input in yours seems to be tan(ρ).

tan (0,79) is outside the domain.

Cheers.
Offline Jean Giraud  
#42 Posted : 29 October 2020 20:28:35(UTC)
Jean Giraud


Rank: Advanced Member

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 4,867
Canada

Was thanked: 801 time(s) in 635 post(s)
... your new algo is freak for the low 'r' region.
Better sanity check wrt the reference log plot.
Project Hanski Soil Drainage Ng ZedGraph [Complete].sm (324kb) downloaded 4 time(s).
Offline hanskl  
#43 Posted : 29 October 2020 20:41:06(UTC)
hanskl


Rank: Advanced Member

Groups: Registered
Joined: 25/10/2017(UTC)
Posts: 94
Norway

Was thanked: 6 time(s) in 4 post(s)
Originally Posted by: Jean Giraud Go to Quoted Post
... your new algo is freak for the low 'r' region.
Better sanity check wrt the reference log plot.
Project Hanski Soil Drainage Ng ZedGraph [Complete].sm (324kb) downloaded 4 time(s).


It seems to plot fine in my last work sheet?
Offline Jean Giraud  
#44 Posted : 30 October 2020 01:59:54(UTC)
Jean Giraud


Rank: Advanced Member

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 4,867
Canada

Was thanked: 801 time(s) in 635 post(s)
All those designed algo fit are interesting, but not so much.
Those reference plots must exist as data sets, matrix table.
At this point, if you consult and want your design pass QA [Quality Assurance],
you can bilinear over the reference matrix.
Bilinear code is universal and will pass QA.
Attach the reference standard matrix, you will get the suite/end of the project.
We have plenty of examples bilinear
Offline hanskl  
#45 Posted : 30 October 2020 03:47:38(UTC)
hanskl


Rank: Advanced Member

Groups: Registered
Joined: 25/10/2017(UTC)
Posts: 94
Norway

Was thanked: 6 time(s) in 4 post(s)
Originally Posted by: Jean Giraud Go to Quoted Post
All those designed algo fit are interesting, but not so much.
Those reference plots must exist as data sets, matrix table.
At this point, if you consult and want your design pass QA [Quality Assurance],
you can bilinear over the reference matrix.
Bilinear code is universal and will pass QA.
Attach the reference standard matrix, you will get the suite/end of the project.
We have plenty of examples bilinear


I made another version of the function that creates a matrix that calculates N.y for a set range of rho, which discards every N.i < N.i-1, and then uses linear interpolation... Seemed to maybe fit the orginal plot the best, but it took way too long to compute, so it was impractical...


If only I had the original data set.. or equations.. or whatever is behind that plot. I have reached out to the author with no luck..


Cheers.

Project_linterp.sm (694kb) downloaded 6 time(s).

Project_linterp.PNG
Offline Jean Giraud  
#46 Posted : 30 October 2020 03:51:54(UTC)
Jean Giraud


Rank: Advanced Member

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 4,867
Canada

Was thanked: 801 time(s) in 635 post(s)
... bilinear demo added on your complete rho °.
Adapt to your reference standard data table, or attach that data table.

Project Hanski Soil Drainage Ng ZedGraph [Complete].sm (343kb) downloaded 4 time(s).
Offline Jean Giraud  
#47 Posted : 30 October 2020 04:08:10(UTC)
Jean Giraud


Rank: Advanced Member

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 4,867
Canada

Was thanked: 801 time(s) in 635 post(s)
Originally Posted by: hanskl Go to Quoted Post
If only I had the original data set.. or equations.. or whatever is behind that plot. I have reached out to the author with no luck..

That is dramatic having a national standard out of the blue.
Those data must be in a repository somewhere.
Those data can be tabulated from the non-log reference plot.
I would use Tracer [freeware], at least ½ a day, no sweat.
Jean
Offline hanskl  
#48 Posted : 30 October 2020 04:17:43(UTC)
hanskl


Rank: Advanced Member

Groups: Registered
Joined: 25/10/2017(UTC)
Posts: 94
Norway

Was thanked: 6 time(s) in 4 post(s)
Originally Posted by: Jean Giraud Go to Quoted Post

That is dramatic having a national standard out of the blue.
Those data must be in a repository somewhere.
Those data can be tabulated from the non-log reference plot.
I would use Tracer [freeware], at least ½ a day, no sweat.
Jean


Theres for sure data behind the plot. The plots are published by a leading university in the 1980s I think, but the users are not presented the underlying data.

For practical purposes, I will use the function with a modified value for c. That seems to be a sweet spot between acceptable accuracy and speed of computation.

Thanks again for all your interest and help with my project, greatly appreciated.

Cheers.
Offline Jean Giraud  
#49 Posted : 30 October 2020 16:17:23(UTC)
Jean Giraud


Rank: Advanced Member

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 4,867
Canada

Was thanked: 801 time(s) in 635 post(s)
Originally Posted by: hanskl Go to Quoted Post
Theres for sure data behind the plot. The plots are published by a leading university in the 1980s I think, but the users are not presented the underlying data.

For practical purposes, I will use the function with a modified value for c. That seems to be a sweet spot between acceptable accuracy and speed of computation.

That's all what you need to carry the project, interpolate at will.
For the plots, ∠(°Wink must be meshed finer ... next today.
ZedGraph has no interest, will be ignored.
Speed of calculation is of no concern for in-situ user.
A most interesting project, nearing end. Some Mathsoft Collaboratory lasted months !

Cheers ... Jean, take care, getting cold here as you guess !

Project Hanski Soil Drainage Ng [Complete Bilinear].sm (31kb) downloaded 11 time(s).
Offline Jean Giraud  
#50 Posted : 31 October 2020 03:29:03(UTC)
Jean Giraud


Rank: Advanced Member

Groups: Registered
Joined: 04/07/2015(UTC)
Posts: 4,867
Canada

Was thanked: 801 time(s) in 635 post(s)
... if you modify your algo N.γ(r,∠°Wink, the remaining of the document
should follow as long as ∠° does not change. Bilinear added.
Jean

Project Hanski Soil Drainage Ng [Complete Bilinear].sm (58kb) downloaded 6 time(s).


Offline hanskl  
#51 Posted : 02 November 2020 00:47:33(UTC)
hanskl


Rank: Advanced Member

Groups: Registered
Joined: 25/10/2017(UTC)
Posts: 94
Norway

Was thanked: 6 time(s) in 4 post(s)
Originally Posted by: Jean Giraud Go to Quoted Post
... if you modify your algo N.γ(r,∠°Wink, the remaining of the document
should follow as long as ∠° does not change. Bilinear added.
Jean




The bilinear was interessting, will have to spend some time to understand the setup.

It's getting pretty cold and dark in Norway too.. and slippery roads!

Cheers.

Edited by user 02 November 2020 00:49:52(UTC)  | Reason: Not specified

Users browsing this topic
3 Pages<123
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.