News

New book: "Bayesian Regression Modeling with INLA"

posted 4 Jan 2018, 08:38 by Havard Rue   [ updated 4 Jan 2018, 08:39 ]

Congratulations to Xiaofeng Wang, Yu Yue Ryan and Julian J Faraway,  for their new book "Bayesian Regression
Modelling with INLA", which is announced on Amazon and ready for preorder (at the time of writing).



"A gentle INLA tutorial"....

posted 29 Dec 2017, 07:28 by Havard Rue   [ updated 29 Dec 2017, 07:32 ]

 has written a gentle INLA tutorial worth reading, here is the link.

H

Introduction to R-INLA for Bayesian Spatial Modeling, Rome 13-15Dec

posted 10 Dec 2017, 02:56 by Havard Rue

Sapienza University, Rome, 13-15 Dec, by Haakon Bakka.  See here for the official announcement.


A 5-day workshop on fitting geostatistical and spatial point process models to spatial survey data (St Andrews, Scotland)

posted 7 Nov 2017, 04:50 by Havard Rue   [ updated 7 Nov 2017, 04:53 ]

26-30th March, 2018 – CREEM, St Andrews

``The Centre for Research into Ecological and Environmental Modelling at the University of St Andrews, in collaboration with the University of Edinburgh, will be holding a 5-day workshop on fitting geostatistical and spatial point process models to spatial survey data. The workshop will be held at the custom workshop facility at the Centre for Research into Ecological and Environmental Modelling in St Andrews, Scotland.

It will cover methods for fitting geostatistical and spatial point process models to data obtained from surveys on which the whole region or population of interest is observed, as well as surveys on which observations are available only in a spatial sample from the region of interest, and surveys on which members of the population of interest are missed with some unknown probability. Inference for this last class of surveys will focus on distance sampling surveys. The workshop covers methods for spatial modelling problems in general, although the examples and exercises in the workshop will focus on ecological surveys.''

MORE INFORMATION HERE



Model criticism and conflict diagnostics using R-INLA

posted 1 Nov 2017, 00:33 by Havard Rue   [ updated 2 Nov 2017, 07:17 by Egil Ferkingstad ]

The code and datafiles required  to reproduce the examples in this paper (recently published in Stat), by Ferkingstad et al (2017), can be found through this link.

H

SHORT COURSE: Spatial and Spatio-Temporal Bayesian Models with R-INLA 16-18 Nov

posted 23 Oct 2017, 23:44 by Havard Rue

Dr Marta Blangairdo will be leading a short course on Spatial and Spatio-Temporal Bayesian Models with R-INLA on the 16th, 17th and 18th of November 2017 following the Big Data, Small Area Symposium. This course will take place at Imperial College London, St Marys Campus, Paddington, London.


Some changes in the R-INLA package

posted 18 Oct 2017, 04:39 by Havard Rue

There are some changes in the testing version of the R-INLA package from version 'INLA_17.10.18', that requires an 
official announcement

1.  The `internal separator' for sections in the Model.ini-file has changed from [NAME] to $NAME$ to allow for names including '[' and ']', like
this 

 Fixed effects:
               mean     sd 0.025quant 0.5quant 0.975quant    mode kld
(Intercept) -9.0750 0.0564    -9.1874  -9.0743    -8.9661 -9.0731   0
x]1-1.5]    -0.0558 0.0890    -0.2314  -0.0555     0.1180 -0.0549   0
x]1.5-3]     0.5655 0.0837     0.4009   0.5656     0.7292  0.5659   0
x]3-5]       0.7962 0.0727     0.6540   0.7960     0.9394  0.7956   0
x]5- ~]      0.9411 0.0771     0.7899   0.9410     1.0925  0.9408   0

This means that if you're running INLA remotely, the local and remote version MUST (for the time being), either both be the stable version,
or the testing version. Otherwise, you'll get an error like this:

> r = inla(y~1, data = data.frame(y=0), inla.call = "remote")
Run /usr/local/bin/inla at hrue@inla.math.ntnu.no

*** ERROR *** inla_build: no section of type = [PROBLEM]



2.  the 32bit versions of the inla-program has been removed. IF there are some that MUST have those, please let us know at help@r-inla.org.
This because it was getting painful to maintain those and 32bit computer are quite rare these days. 

Best
H




Fractional Gaussian Noise models, are now available!

posted 28 Sep 2017, 00:00 by Havard Rue

Due to new results, we have added support for FGN models in R-INLA. The report is here 

H


Want to do a PhD?

posted 27 Sep 2017, 23:53 by Havard Rue   [ updated 28 Sep 2017, 00:04 ]

I have an open spot in my research group ,  with application deadline 15th Oct.  It is well funded. 
Please contact me to discuss options.

Håvard Rue

Check out kaust.edu.sa

Changes in the build and web-hosting of R-INLA

posted 21 May 2017, 03:07 by Havard Rue

We are migrating away from www.math.ntnu.no/inla and have created a new domain
to host the prebuildt packages;

    inla.r-inla-download.org

which will also keep old packages.  There are still some links pointing to
www.math.ntnu.no/inla but we will fix that later this year.

We have changed the versioning system into

    INLA_YEAR.MONTH.DAY[-MINOR]

each with two digits.The version INLA_17.05.21  is the version from 21st May 2017, and
you can find the tag 'Version_17.05.21' in the code repository. This should be
more intuitive than the old system. We'll update the stable package in the near future.

The Linux binaries are now built with CentOS-7  and we use R-3.4 for Linux/Mac/Win.

If you experience any issues due to these changes, please let us know.

1-10 of 113