Menu
▾
▴
kriging-help — Public mailing list where STK users can ask for help
You can subscribe to this list here.
| 2013 |
Jan
|
Feb
(2) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2014 |
Jan
(5) |
Feb
|
Mar
|
Apr
(2) |
May
|
Jun
|
Jul
(5) |
Aug
|
Sep
(2) |
Oct
|
Nov
|
Dec
(3) |
| 2015 |
Jan
|
Feb
|
Mar
(2) |
Apr
|
May
(4) |
Jun
(5) |
Jul
|
Aug
|
Sep
(6) |
Oct
|
Nov
|
Dec
|
| 2016 |
Jan
|
Feb
|
Mar
(10) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(3) |
Dec
|
| 2017 |
Jan
|
Feb
|
Mar
(2) |
Apr
(4) |
May
|
Jun
(3) |
Jul
(9) |
Aug
|
Sep
|
Oct
(6) |
Nov
|
Dec
(1) |
| 2018 |
Jan
(1) |
Feb
(1) |
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2019 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
(1) |
| 2020 |
Jan
|
Feb
|
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(1) |
Oct
|
Nov
|
Dec
|
| 2022 |
Jan
(2) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Julien B. <jul...@ce...> - 2022-01-31 22:12:38
|
Le 31/01/2022 à 21:20, Seth Cousin a écrit : > > I work in the aerospace world, specifically doing calibration and > characterization of detectors and space instruments (cubesats, space > spectrographic satellites etc.) We always need to interpolate our > datasets to fully characterize an instrument i.e. a satellite may need > to view light from 300nm to 2300nm. In the lab we can use a laser at a > few wavelengths, but then rely on interpolation to complete our > calibrations. Kriging seems like a great tool for this as ultimately > the calibration is defined by its uncertainty, and the variances > computed via kriging can be used to calculate them. > > I have one specific question, that may not be a typical question for > this tool. Is there a way to set any limits on a krig interpolation? > The exact problem I have is that I have some recorded line shapes, > that I fit some distribution to. I am then using kriging to > interpolate the coefficients of the fit. Some of the coefficients > cannot go below zero (it makes no physical sense i.e. Gaussian width > cannot be<0), so I am trying to see if there is a way to limit the > kriged value to > 0. Hi Seth, One possible, simple approach is to transform the data to get rid of the positivity constraint. For instance, you can take the log of your positive data, and then map the kriging predictions and confidence intervals back to the original domain using the exponential. (Other mappings may turn out to be more suitable to your particular case.) Another approach is to used some form of "constrained kriging" [a,b] but this is not readily available in the STK toolbox at the moment. HTH, feel free to ask for further details. @++ Julien References: [a] S. Da Veiga & A. Marrel (2012), Gaussian process modeling with inequality constraints <http://www.numdam.org/item/AFST_2012_6_21_3_529_0/>, Annales de la faculté des sciences de Toulouse Mathématiques, 21(3), 529-555. [b] H. Maatouk & X. Bay (2017) Gaussian process emulators for computer experiments with inequality constraints <https://link.springer.com/article/10.1007/s11004-017-9673-2>, Mathematical Geosciences, 49, 557–582. |
|
From: Seth C. <Set...@la...> - 2022-01-31 20:21:07
|
I work in the aerospace world, specifically doing calibration and characterization of detectors and space instruments (cubesats, space spectrographic satellites etc.) We always need to interpolate our datasets to fully characterize an instrument i.e. a satellite may need to view light from 300nm to 2300nm. In the lab we can use a laser at a few wavelengths, but then rely on interpolation to complete our calibrations. Kriging seems like a great tool for this as ultimately the calibration is defined by its uncertainty, and the variances computed via kriging can be used to calculate them. I have one specific question, that may not be a typical question for this tool. Is there a way to set any limits on a krig interpolation? The exact problem I have is that I have some recorded line shapes, that I fit some distribution to. I am then using kriging to interpolate the coefficients of the fit. Some of the coefficients cannot go below zero (it makes no physical sense i.e. Gaussian width cannot be<0), so I am trying to see if there is a way to limit the kriged value to > 0. |
|
From: Julien B. <jul...@ce...> - 2020-09-09 12:23:39
|
Hi all, The bug tracker on Sourceforge is no longer actively used by the STK project. Please create your tickets for bug reports & feature request here instead: https://github.com/stk-kriging/stk/issues. @++ Julien |
|
From: Julien B. <jul...@ce...> - 2020-03-23 15:44:39
|
Le 23/03/2020 à 14:57, Francesco Potortì a écrit : >> linear, cubic, and natural methods are all triangulation based. they take >> the scattered points, use octave's delauney triangulation, then calculate >> local interpolants within the triangles. linear (actually bilinear i guess) >> is simply a weighted some of triangle areas (the interpolation point >> creates three sub-triangles within the mesh triangle, and value is a >> weighted projection inversely proportional to sub-triangle size, where a >> big sub-triangle has you more strongly weighted toward the point not >> associated with that sub-triangle). the v4 method does not triangulate, >> and does a weighted interpolation over the entire mesh. that can get slow >> and memory intensive, but it's currently the only 'smooth' interpolation >> function Octave has implemented (sort of). > Have you considered Kriging? I have used the stk package from > Octave-forge with profit. Happy to read that ;-) I'm sorry I haven't been following this thread carefully, but if anybody needs help to perform scattered-data interpolation with stk, feel free to ask here or on kri...@li.... @++ Julien |
|
From: Julien B. <jul...@ce...> - 2019-12-04 13:15:27
|
Le 28/11/2019 à 15:56, Jonas Allgeier a écrit : > I just started to work with your project and noticed the following > behavior with stk_example_kb03: If I increase NI (the number of > observations) to 200 and beyond, I get the warning message "Matrix is > close to singular or badly scaled. Results may be inaccurate". This is > not intuitive to me, since I'd expect that with an increasing number > of known data points it must be easier to approximate the unknown > field. However, I'm very new to Kriging and I might misunderstand the > underlying math. Could you help me explain why I get that message, if > it is of relevance (i.e., is the result really inaccurate?) and how I > could prevent it? Hi Jonas, Yes, this is expected if you use a "noiseless" kriging model (which is the case in this example, check that model.lognoisevariance is -inf). The reason is that the covariance matrix becomes ill-conditioned when n -> inf, since points are getting closer and closer to each others. The warning in itself doesn't mean that things are going wrong, you can check that the result is still OK with NI = 300 or 400 (look at the cross-validation plots for instance). The easiest and usual fix, to get rid of these warnings and avoid numerical problems, is to add just a little bit of "noise" to your model, which will strengthen the diagonal of the covariance matrix. Look at the changes that I just made to stk_example_kb03 : https://sourceforge.net/p/kriging/hg/ci/1020955db5f998bc277463a683116588871d612f/ HTH, feel free to ask further questions. @++ Julien |
|
From: Jonas A. <jon...@un...> - 2019-11-28 15:13:29
|
Dear developer(s) of the STK toolbox, I just started to work with your project and noticed the following behavior with stk_example_kb03: If I increase NI (the number of observations) to 200 and beyond, I get the warning message "Matrix is close to singular or badly scaled. Results may be inaccurate". This is not intuitive to me, since I'd expect that with an increasing number of known data points it must be easier to approximate the unknown field. However, I'm very new to Kriging and I might misunderstand the underlying math. Could you help me explain why I get that message, if it is of relevance (i.e., is the result really inaccurate?) and how I could prevent it? Kind regards, Jonas |
|
From: Julien B. <jb...@us...> - 2018-07-03 06:24:00
|
Le 03/07/2018 à 07:39, sa...@us... a écrit : > Hello dear Julien > > I want to create a kriging model from my experimental data. How can I > begin with STK Toolbox ? I neea an example which explaining the > importing data to the program and step-by-step procedure of creating a > kriging model. > > I can’t thank you enough > > with best regards, > Saeed Farahi Shahri > Dear Saeed, You should read the example scripts in the "example" directory to learn the basics of STK. Basically, assuming that you want a noisy kriging model, it will look like this : ----- % load your data into a n x p array of input points x_obs, and an n x 1 array of responses z_obs M_prior = stk_model (@stk_materncov52); M_prior.lnv = nan; % if you want a noisy kriging model [M_prior.param, M_prior.lnv] = stk_param_estim (M_prior, x_obs, z_obs); z_pred = stk_predict (M_prior, x_obs, z_obs, x_pred); ----- If you're going to use the same model several times to make predictions, you will prefer the following variation : ----- M_prior = stk_model (@stk_materncov52); M_prior.lnv = nan; % if you want a noisy kriging model [M_prior.param, M_prior.lnv] = stk_param_estim (M_prior, x_obs, z_obs); M_post = stk_model_gpposterior (M_prior, x_obs, z_obs); z_pred1 = stk_predict (M_post, x_pred1); ... z_pred2 = stk_predict (M_post, x_pred2); ... ----- If you need a more explicit example, including how to load your data, plot the results, etc. please send some data. HTH, Julien. |
|
From: Julien B. <jul...@ce...> - 2018-03-01 07:31:44
|
Hi again, kriging-help list ! Please ignore this message as well, this is just another test... @++ Julien |
|
From: Julien B. <jul...@ce...> - 2018-02-28 05:29:25
|
Hi kriging-help list ! Please ignore this message, this is just a test... @++ Julien |
|
From: Julien B. <jul...@ce...> - 2018-01-08 13:00:23
|
Le 18/12/2017 à 07:38, Julien Bect a écrit : > Le 15/12/2017 à 16:18, Derks, Eduard a écrit : >> >> Dear Julien >> >> When I try to run the first STK example I get an error that the field >> ‘colnames’ is not known. >> >> Error using stk_hrect/get (line 46) >> >> There is no field named colnames >> >> Error in stk_sampling_regulargrid (line 67) >> >> box_colnames = get (box, 'colnames'); >> >> Error in stk_dataset_twobumps (line 51) >> >> ref.xt = stk_sampling_regulargrid (400, [], ref.box); >> >> In installed the toolbox in >> >> MATLAB Version: 9.0.0.341360 (R2016a) >> >> I’ld greatly appreciate your help. >> >> Kind regards, >> >> Eduard >> > > Hello Eduard, > > Which version of stk are you using ? On which OS ? > > Is this reproducible (have you tried to close/restart Matlab several > times and seen the problem each time) ? > > Could you set a break point (dbstop if error) and save the "box" > object for inspection ? Hello Eduard, I have tried to run the first example of STK 2.5.0 in Matlab R2016a on Windows 10. The example worked fine and I could not reproduce the error that you report. Please provide more detailed information if you still need help. @++ Julien |
|
From: Julien B. <jul...@ce...> - 2017-12-18 06:38:35
|
Hello Eduard, Which version of stk are you using ? On which OS ? Is this reproducible (have you tried to close/restart Matlab several times and seen the problem each time) ? Could you set a break point (dbstop if error) and save the "box" object for inspection ? @++ Julien Le 15/12/2017 à 16:18, Derks, Eduard a écrit : > > Dear Julien > > When I try to run the first STK example I get an error that the field > ‘colnames’ is not known. > > Error using stk_hrect/get (line 46) > > There is no field named colnames > > Error in stk_sampling_regulargrid (line 67) > > box_colnames = get (box, 'colnames'); > > Error in stk_dataset_twobumps (line 51) > > ref.xt = stk_sampling_regulargrid (400, [], ref.box); > > In installed the toolbox in > > MATLAB Version: 9.0.0.341360 (R2016a) > > I’ld greatly appreciate your help. > > Kind regards, > > Eduard > |
|
From: Julien B. <jul...@ce...> - 2017-10-30 15:12:13
|
Le 26/10/2017 à 17:44, Julien Bect a écrit :
> Le 26/10/2017 à 17:42, Julien Bect a écrit :
>> Le 26/10/2017 à 14:23, Rackl Michael a écrit :
>>> Thank you for pointing out that there is no particular best
>>> practice. Also: I did not know the field “data” existed for
>>> stk_dataframe objects.
>>>
>>> It would be nice if this was mentioned in the documentation, since
>>> my current workaround was
>>> x = struct2cell(x); % followed by
>>> x = x{1,1};
>>>
>>> which was unnecessarily complicated.
>>
>> You're perfectly right, this should documented.
>>
>> I will document this in the help of stk_dataframe, this will be
>> available in the next release.
Here it is :
https://sourceforge.net/p/kriging/hg/ci/af474083e8212e7a0bbdf74cf736d5496352c5f2/
@++
Julien
|
|
From: Julien B. <jul...@ce...> - 2017-10-26 15:44:14
|
Le 26/10/2017 à 17:42, Julien Bect a écrit :
> Le 26/10/2017 à 14:23, Rackl Michael a écrit :
>> Thank you for pointing out that there is no particular best practice.
>> Also: I did not know the field “data” existed for stk_dataframe objects.
>>
>> It would be nice if this was mentioned in the documentation, since my
>> current workaround was
>> x = struct2cell(x); % followed by
>> x = x{1,1};
>>
>> which was unnecessarily complicated.
>
> You're perfectly right, this should documented.
>
> I will document this in the help of stk_dataframe, this will be
> available in the next release.
>
> If you want a more verbose display of stk_dataframe objects, you can
> also try this :
>
> stk_options_set ('stk_dataframe', 'disp_format', 'verbose');
>
> (There is a tiny display bug that I just noticed when testing this for
> you, that I will also fix for the next release.)
Just one more remark : the good thing with "double (x)" is that it works
for many types of inputs (single, double, int32, stk_dataframe, etc.)
OTOH, "x.data" works for stk_dataframe only.
|
|
From: Julien B. <jul...@ce...> - 2017-10-26 15:42:58
|
Le 26/10/2017 à 14:23, Rackl Michael a écrit :
> Thank you for pointing out that there is no particular best practice.
> Also: I did not know the field “data” existed for stk_dataframe objects.
>
> It would be nice if this was mentioned in the documentation, since my
> current workaround was
> x = struct2cell(x); % followed by
> x = x{1,1};
>
> which was unnecessarily complicated.
You're perfectly right, this should documented.
I will document this in the help of stk_dataframe, this will be
available in the next release.
If you want a more verbose display of stk_dataframe objects, you can
also try this :
stk_options_set ('stk_dataframe', 'disp_format', 'verbose');
(There is a tiny display bug that I just noticed when testing this for
you, that I will also fix for the next release.)
@++
Julien
|
|
From: Rackl M. <Ra...@fm...> - 2017-10-26 12:23:29
|
Hi Julien,
Thank you for pointing out that there is no particular best practice. Also: I did not know the field "data" existed for stk_dataframe objects.
It would be nice if this was mentioned in the documentation, since my current workaround was
x = struct2cell(x); % followed by
x = x{1,1};
which was unnecessarily complicated.
Best regards,
Michael
|
|
From: Julien B. <jul...@ce...> - 2017-10-20 20:23:20
|
Le 20/10/2017 à 11:06, Rackl Michael a écrit : > > Dear Julien, > > the news for STK update 2.5.0 say: > > “Support old-style STK structures (with a .a field) has been removed.” > > Is there now another way to extract array-like information from STK > structures? > > Best regards, > > Michael > Hi Michael, Sure, there are other ways ;-) If x is an stk_dataframe object, you can extract the underlying numerical data using "double (x)" or "x.data". Note that many frequent operations are already overloaded for stk_dataframe objects, which means that you shouldn't have to do that very often. Don't hesitate to ask if you feel that more functions should be overloaded for stk_dataframe objects. @++ Julien |
|
From: Rackl M. <Ra...@fm...> - 2017-10-20 09:21:20
|
Dear Julien, the news for STK update 2.5.0 say: "Support old-style STK structures (with a .a field) has been removed." Is there now another way to extract array-like information from STK structures? Best regards, Michael |
|
From: Rackl M. <Ra...@fm...> - 2017-07-21 08:17:09
|
Le 20/07/2017 à 15:33, Rackl Michael a écrit : is there a function in the STK package, which can be used to carry out sequential optimization based on expected improvement (EI)? I found the functions "stk_minimize_unconstrained" and "stk_minimize_constrainted", but I am not sure what they do. Hi Michael, Assuming that you have stk 2.4.2, the computation of the EI criterion is done by stk_sampcrit_ei_eval (): http://kriging.sourceforge.net/htmldoc/function/stk_sampcrit_ei_eval.html You have a simple demo of EI-based optimization in 1D in stk_example_doe03: http://kriging.sourceforge.net/htmldoc/function/stk_example_doe03.html This demo uses a fixed grid. You can easily adapt it to 2D or 3D problems, if you don't need very accurate results. If you want to address higher-dimensional problems, or get more precise convergence, you will need to work a little harder... HTH, Julien. PS : the stk_minimize_* functions are not what you're looking for Hi Julien, thank you for your answer! best regards, Michael |
|
From: Julien B. <jul...@ce...> - 2017-07-21 06:38:56
|
Le 20/07/2017 à 15:33, Rackl Michael a écrit : > is there a function in the STK package, which can be used to carry out > sequential optimization based on expected improvement (EI)? > > I found the functions “stk_minimize_unconstrained” and > “stk_minimize_constrainted”, but I am not sure what they do. > Hi Michael, Assuming that you have stk 2.4.2, the computation of the EI criterion is done by stk_sampcrit_ei_eval (): http://kriging.sourceforge.net/htmldoc/function/stk_sampcrit_ei_eval.html You have a simple demo of EI-based optimization in 1D in stk_example_doe03: http://kriging.sourceforge.net/htmldoc/function/stk_example_doe03.html This demo uses a fixed grid. You can easily adapt it to 2D or 3D problems, if you don't need very accurate results. If you want to address higher-dimensional problems, or get more precise convergence, you will need to work a little harder... HTH, Julien. PS : the stk_minimize_* functions are not what you're looking for |
|
From: Rackl M. <Ra...@fm...> - 2017-07-20 13:51:22
|
Dear members of the STK mailing list, is there a function in the STK package, which can be used to carry out sequential optimization based on expected improvement (EI)? I found the functions "stk_minimize_unconstrained" and "stk_minimize_constrainted", but I am not sure what they do. Best regards, Michael |
|
From: Julien B. <jul...@ce...> - 2017-07-16 07:50:36
|
Le 16/07/2017 à 00:51, Daniel Oliveira a écrit : > > I installed STK manually because Octave return an error when I use > 'pkg install -forge stk'. Probably is a problem with curl. > > I installed the latest release of STK (2.4.2) and the problem is gone. > > Thank you very much. Glad to hear that the problem is gone, although I don't exactly know what it was... @++ Julien |
|
From: Daniel O. <dan...@ho...> - 2017-07-15 22:52:02
|
You are right. The Octave version is 4.2.1. I installed STK manually because Octave return an error when I use 'pkg install -forge stk'. Probably is a problem with curl. I installed the latest release of STK (2.4.2) and the problem is gone. Thank you very much. Daniel Oliveira Whatsapp: +55 31 997 07 207 Skype: dan...@ho... linkedin: https://www.linkedin.com/in/daniel-borges-de-oliveira-656a2571 ________________________________ From: Julien Bect <jul...@ce...> Sent: Friday, July 14, 2017 6:10 PM To: Daniel Oliveira Cc: emm...@ce...; kri...@li... Subject: Re: Problem example stk_example_kb01.m - STK 2.4.0 Le 14/07/2017 à 20:36, Daniel Oliveira a écrit : Below is the configuration that I used. Operational system: Debian GNU/Linux 8. Compiled Octave 4.1.2 with this flags: I assume you mean 4.2.1 ? CFLAGS="-I/opt/suitesparse-4.5.5/include" \ LDFLAGS="-L/opt/openblas/lib/ -L/opt/suitesparse-4.5.5/lib" \ ./configure \ --prefix=/opt/octave-4.2.1 \ --with-shell=/bin/bash \ --with-hdf5-libdir=/opt/hdf5-1.8.14/lib \ --with-hdf5-includedir=/opt/hdf5-1.8.14/include \ --with-curl-includedir=/usr/include/curl \ --with-curl-libdir=/usr/lib/x86_64-linux-gnu \ --with-glpk-includedir=/opt/glpk-4.61/include \ --with-glpk-libdir=/opt/glpk-4.61/lib/ \ --with-java-includedir=/opt/java/jdk1.8.0_131/include \ --with-java-libdir=/opt/java/jdk1.8.0_131/lib \ --with-java-homedir=/opt/java/jdk1.8.0_131 \ --with-cxsparse-includedir=/opt/suitesparse-4.5.5/include \ --with-cxsparse-libdir=/opt/suitesparse-4.5.5/lib Commands used to install stk: # wget https://ufpr.dl.sourceforge.net/project/octave/Octave%20Forge%20Packages/Individual%20Package%20Releases/stk-2.4.0.tar.gz && octave-cli --eval 'pkg install stk-2.4.0.tar.gz' I used the command "pkg load stk" inside Octave command window to load stk. Please try again with the latest release of STK (2.4.2) and see if the problem persists. By the way, you don't need to download it manually, you just have to do : pkg install -forge stk |
|
From: Julien B. <jul...@ce...> - 2017-07-15 07:40:50
|
Le 14/07/2017 à 23:10, Julien Bect a écrit : >> Commands used to install stk: >> # wget >> https://ufpr.dl.sourceforge.net/project/octave/Octave%20Forge%20Packages/Individual%20Package%20Releases/stk-2.4.0.tar.gz >> && octave-cli --eval 'pkg install stk-2.4.0.tar.gz' >> >> I used the command "pkg load stk" inside Octave command window to >> load stk. > > Please try again with the latest release of STK (2.4.2) and see if the > problem persists. > > By the way, you don't need to download it manually, you just have to do : > > pkg install -forge stk I have tried to install stk 2.4.0 on Debian 9 (stretch) exactly as you did (wget + octave-cli --eval ..."), both on Octave 4.0.3 (from Debian repos) and Octave 4.2.1+ (tip of stable branch, compiled from source). Unfortunately, I could not reproduce the problem that you describe. Does it occur reproducibly each time you do 1) start Octave 4.2.1, 2) pkg load stk and then 3) stk_example_kb01 ? |
|
From: Julien B. <jul...@ce...> - 2017-07-14 21:10:19
|
Le 14/07/2017 à 20:36, Daniel Oliveira a écrit : > > Below is the configuration that I used. > > Operational system: Debian GNU/Linux 8. > > Compiled Octave 4.1.2 with this flags: I assume you mean 4.2.1 ? > CFLAGS="-I/opt/suitesparse-4.5.5/include" \ > LDFLAGS="-L/opt/openblas/lib/ -L/opt/suitesparse-4.5.5/lib" \ > ./configure \ > --prefix=/opt/octave-4.2.1 \ > --with-shell=/bin/bash \ > --with-hdf5-libdir=/opt/hdf5-1.8.14/lib \ > --with-hdf5-includedir=/opt/hdf5-1.8.14/include \ > --with-curl-includedir=/usr/include/curl \ > --with-curl-libdir=/usr/lib/x86_64-linux-gnu \ > --with-glpk-includedir=/opt/glpk-4.61/include \ > --with-glpk-libdir=/opt/glpk-4.61/lib/ \ > --with-java-includedir=/opt/java/jdk1.8.0_131/include \ > --with-java-libdir=/opt/java/jdk1.8.0_131/lib \ > --with-java-homedir=/opt/java/jdk1.8.0_131 \ > --with-cxsparse-includedir=/opt/suitesparse-4.5.5/include \ > --with-cxsparse-libdir=/opt/suitesparse-4.5.5/lib > > Commands used to install stk: > # wget > https://ufpr.dl.sourceforge.net/project/octave/Octave%20Forge%20Packages/Individual%20Package%20Releases/stk-2.4.0.tar.gz > && octave-cli --eval 'pkg install stk-2.4.0.tar.gz' > > I used the command "pkg load stk" inside Octave command window to load > stk. Please try again with the latest release of STK (2.4.2) and see if the problem persists. By the way, you don't need to download it manually, you just have to do : pkg install -forge stk |
|
From: Daniel O. <dan...@ho...> - 2017-07-14 18:36:36
|
Thank you for reply. Below is the configuration that I used. Operational system: Debian GNU/Linux 8. Compiled Octave 4.1.2 with this flags: CFLAGS="-I/opt/suitesparse-4.5.5/include" \ LDFLAGS="-L/opt/openblas/lib/ -L/opt/suitesparse-4.5.5/lib" \ ./configure \ --prefix=/opt/octave-4.2.1 \ --with-shell=/bin/bash \ --with-hdf5-libdir=/opt/hdf5-1.8.14/lib \ --with-hdf5-includedir=/opt/hdf5-1.8.14/include \ --with-curl-includedir=/usr/include/curl \ --with-curl-libdir=/usr/lib/x86_64-linux-gnu \ --with-glpk-includedir=/opt/glpk-4.61/include \ --with-glpk-libdir=/opt/glpk-4.61/lib/ \ --with-java-includedir=/opt/java/jdk1.8.0_131/include \ --with-java-libdir=/opt/java/jdk1.8.0_131/lib \ --with-java-homedir=/opt/java/jdk1.8.0_131 \ --with-cxsparse-includedir=/opt/suitesparse-4.5.5/include \ --with-cxsparse-libdir=/opt/suitesparse-4.5.5/lib Commands used to install stk: # wget https://ufpr.dl.sourceforge.net/project/octave/Octave%20Forge%20Packages/Individual%20Package%20Releases/stk-2.4.0.tar.gz && octave-cli --eval 'pkg install stk-2.4.0.tar.gz' I used the command "pkg load stk" inside Octave command window to load stk. Daniel Oliveira Whatsapp: +55 31 997 07 207 Skype: dan...@ho... linkedin: https://www.linkedin.com/in/daniel-borges-de-oliveira-656a2571 ________________________________ From: Julien Bect <jul...@ce...> Sent: Friday, July 14, 2017 2:47 PM To: Daniel Oliveira Cc: emm...@ce...; kri...@li... Subject: Re: Problem example stk_example_kb01.m - STK 2.4.0 Le 14/07/2017 à 18:33, Daniel Oliveira a écrit : Hello guys, I am studying kriging throught Octave STK 2.4.0 package. When I execute the example "examples / 01_kriging_basics / stk_example_kb01.m" I recive this error: Error: There is no field named data Error: called from Stk_error at line 80 column 1 Get at line 46 column 13 Subsection 51 column 11 Stk_sampling_regulargrid at line 81 column 6 Stk_dataset_twobumps at line 51 column 8 Stk_example_kb01 at line 52 column 12 Can you help me solve this or indicate what is the problem? Hi Daniel, This vaguely rings a bell. Probably an Octave problem. Which version of Octave are you using ? On which OS ? How exactly did you install and load STK ? @++ Julien |
1 message has been excluded from this view by a project administrator.
