Questions tagged [kriging]
Kriging is spatial prediction based on a stochastic model of a spatial random field. Such models and methods can also be used in non-spatial context, and then often known as gaussian processes.
97
questions
1
vote
0
answers
9
views
Can kriging method be used for dataset that is not spatial?
Sorry if the question sounds stupid. I am new to this. Consider the following dataset:
...
2
votes
0
answers
19
views
Kriging : when it is said kriging is an unbiaised estimator, is that synonymous with saying it is an exact interpolator?
I feel like they are not synonymous, but I cannot intuitively explain the difference between "unbiased" and "exact." In other words, I am asking about the difference between the ...
- 21
0
votes
0
answers
24
views
Covariance estimation in simple kriging: the "dead end" problem (authorization of using variogram)
I am confused about the presentation of using variogram in simple kriging in the book of Multivariate geostatistics by H. Wackernagel (1995).
I understand the process of derivation of simple kriging ...
- 21
0
votes
0
answers
29
views
Building a mixed-categorical-continuous probability distribution
I have a variable space that consists of some mixed categorical, integer, and real valued variables i.e $X = (x_1, ..., x_n), x_1 \in \mathbb{Z}, x_2 \in \mathbb{R}, x_n \in \{1,2\}$, and a (small) ...
- 135
0
votes
0
answers
28
views
How to implement a linear regression (like the SMT toolbox) using sklearn's Gaussian Process?
The SMT package allows a poly option (see the documentation here and source code here), where a functional form (constant, linear, or quadratic) is assumed as a ...
- 127
5
votes
1
answer
114
views
Lipschitzness of posterior mean of Gaussian process?
Let $T$ be a compact set, and let $K \colon T \times T \to \mathbb{R}$ be a positive definite kernel. Consider the canonical pseudo-distance
$$d_K(x,y) = \sqrt{K(x,x) + K(y,y) - 2 K(x,y)}.$$
Let $f$ ...
- 151
1
vote
1
answer
62
views
Declustering spatial data and kriging
I am doing a simple project on a mining dataset and came across some problems:
Should I use the declustered values to model the variogram?
Do I apply ordinary kriging to the original or to the ...
1
vote
0
answers
70
views
Gaussian Process as a time series model, pros and cons
When talkin' about time series models, ARIMA, SARIMAX, Prophet or VAR is what mostly is on the table. But what would the benefit be of using a Gaussian Process Regressor instead, in which situations ...
- 223
0
votes
0
answers
65
views
Posterior covariance of the function values and derivatives of a Gaussian Process
I'm trying to figure out how to calculate the cross-terms of the covariance matrix between the function values and derivatives of a Gaussian Process. For context, this is needed for Gaussian Process ...
- 101
0
votes
0
answers
17
views
Geostatistics with different measurement errors
I'm having a hard time identifying the subject and possible resources of my problem.
I have spatio-temporal data (Z) coming from environmental sensors that travel around a city. We have the spatial ...
- 1
0
votes
0
answers
50
views
What am I not understanding about semivariogram and Normal Score Transformation?
I have generated this two dimensional random field:
This is done following this page. In particular, I have selected t=23 as dataframe and I have changed some parameters.
As you can noticed, I have ...
- 111
1
vote
1
answer
66
views
Generate a syntetic log-normal two dimensional random field
I would like to test some functions that I wrote related to the kriging applied to rain data. In order to do that, I would like to generate a synthetic log-normal 2D random field.
The idea is to ...
- 111
1
vote
0
answers
40
views
How to constrain ordinary kriging weights to sum to 1 in R
I need to perform ordinary kriging on a dataset and I understand that I need the weights to sum to 1, I just don't understand how to set that up properly. For example, my covariance function is C(h) =...
- 11
10
votes
2
answers
1k
views
Estimating probability of attack in Ukraine, given count data
I was looking at some attack count data in Ukraine for different days. The data is gathered from the ACCLED dataset, and there is a picture below. The picture shows individual attacks, but I can apply ...
- 1,472
0
votes
0
answers
28
views
Is there a Sequential Gaussian Simulation that uses Ordinary Kriging?
As far as I have read, Sequential Gaussian Simulation always uses Simple Kriging. Is there any chance that it uses Ordinary Kriging?
- 11
0
votes
1
answer
36
views
Kriging with positivity constraint using the STK toolbox
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 ...
- 1
6
votes
1
answer
313
views
What is the difference between a non-zero nugget and a noise term in Kriging/GPR?
With some Gaussian Process Regression/Kriging models, it's possible to specify both a non-zero nugget, and a noise term. For example, in Scikit-learn's GPR model, there is an ...
- 5,183
1
vote
1
answer
189
views
Analytical Leave-one-out prediction variance for Kriging
I make extensive use of Kriging (Gaussian Process regression) methods in my work especially using the leave-one-out error calculation that you can get from the Gram matrix.
Background:
To compute the ...
- 139
1
vote
1
answer
467
views
How can concentrated (profile) log marginal likelihood be used to optimize the mean and scale(outputscale) parameters in Gaussian Process Regression?
The log marginal likelihood which is used in Gaussian Process Regression comes from a Multivariate Normal pdf Gaussian Processes for Machine Learning, p.19, eqn. 2.30, Surrogates, Chapter 5, eqn. 5.4
\...
- 13
4
votes
1
answer
341
views
Who first proposed Bayesian optimisation with Gaussian processes?
From what I understand, the 'standard' approach to Bayesian Optimisation uses a Gaussian process for the prior (as opposed to more recent proposals like TPE or Bayesian Optimisation with random ...
- 139
1
vote
0
answers
203
views
Interpreting a Distance & Time 3D Variogram for Variogram modeling
I am trying to understand some concepts of variograms. I have made several variogram models in R and am trying to understand exactly what they mean. My data is ...
- 171
1
vote
0
answers
183
views
Variogram fit in R not converging
I have taken a shapefile from Open NYC Data and performed the following method. My end goal is to predict Taxi trip_duration at various points across the city of ...
- 171
1
vote
1
answer
499
views
Cokriging and collocated cokriging data requirements
In this wiki article and elsewhere in educational materials/papers, I have seen people refer to the idea that secondary data, if used (appropriately) in cokriging or collocated cokriging, is usually ...
- 55
3
votes
0
answers
872
views
Differences between Kriging and Gaussian Process Regression
I am having quite difficult time to clearly understand the differences between Kriging and Gaussian Process Regression.
Here is what I have understood so far:
For simple kriging (mean value known), ...
- 311
2
votes
0
answers
98
views
Conditional distribution of Ornstein-Uhlenbeck on two fixed points
The conditional distribution of a Ornstein-Uhlenbeck $X(t)$ conditional on $X(0)$ is given by
$$
X(t)|X(0) = X(0)e^{-t} + \mu(1 - e^{-t})
$$
This process is usually only defined for $t>0$ (future ...
- 546
1
vote
0
answers
74
views
(Co)kriging / co-located kriging with heterogenous measurement errors
I have a large set of polygons on a map, some of which contain data on 2 count variables, say ($z_{1}$ and $z_{2}$) that are correlated. In fact, $z_{2}$ most likely causes $z_{1}$ without the ...
- 55
2
votes
1
answer
121
views
Intuition Behind Correlation Function in Kriging Models
I'm thinking and researching extensively to interpret the parameter $\theta$ (activeness parameter) in Gaussian correlation function in a Kriging model, namely as:
$$
K(h;\theta)=exp(-h^2/(2\theta^2))
...
- 63
1
vote
0
answers
142
views
Intuitive explanation of Gaussian Process Regression
How would you intuitively explain the idea behind Gaussian Process Regression to someone unfamiliar with stochastic processes? Especially the point where you discuss modeling covariance, choice of ...
- 11
1
vote
0
answers
28
views
“Jumping” among several interpolation techniques?
I am comparing several interpolation methods using monthly climatic data, through RMSE and a 10-fold cross-validation scheme.
What I'm observing is that the performances vary from one month to ...
- 23
4
votes
1
answer
880
views
Modeling trends with Gaussian Processes
I am trying to use Gaussian Process Regression to model data that has a clear upward rising trend. Here is a basic plot of the data -
The most well-known example of the flexibility of GPs in such ...
- 1,541
1
vote
0
answers
25
views
learning time series by machine learning
I am trying to learn a mapping between coordinates x,y,z,d, with d=sqrt(x**2 + y**2 + z**2), and a scalar. Each ...
- 163
1
vote
0
answers
78
views
When regularizing based on an informative prior, how to give model a little more freedom to partially reject regulariziation
I am new here I hope this question is appropriate.
I am modelling a spatial domain, whereby I have repeated measures at n locations. I make a bayesian linear model at each n locations based on about ...
- 261
1
vote
1
answer
223
views
Can I replace the distribution in Gaussian Process Regression with a different regression?
Sorry for the confusing title. Let me try to clarify:
I have a time series of wind speeds with some missing points here and there. I want to interpolate these points and have tried mainly polynomial ...
- 143
2
votes
1
answer
489
views
Is the spherical covariance function not positive definite for d > 3?
I read in a textbook (Japanese one) that the spherical covariance function is only valid for dimensions $d = 1,$ $2,$ and $3.$ I have the following questions:
Does that mean the spherical covariance ...
- 23
2
votes
0
answers
20
views
Cokriging variances differ using cross validation
I'm investigating cokriging using various metals in the Meuse dataset but the variances output by R when I predict values at gridded points differ substantially from the variances produced by cross ...
0
votes
0
answers
31
views
Why do you need a variogram for Kriging? goldingn/gpe package?
I am using golingn/gpe (github) package, and it does not provide a variogram and instead look at co-variances.
Is it possible to do kriging without providing variograms?
- 1
0
votes
0
answers
209
views
How to handle multiple points at the same location in spatial interpolation?
I am new to the topic of spatial interpolation and would appreciate your opinion on a general question which has arisen.
Suppose I have a data set containing rental rates for different apartments in ...
- 623
1
vote
1
answer
198
views
Poor performance from `gstat::krige` with a noise predictor
I'm new to kriging, and I'm considering replacing a use of inverse-distance weighting (IDW) in a spatial modeling project (implemented with gstat::idw in R) with a ...
- 19.5k
0
votes
0
answers
65
views
Gaussian process with interval observations
The stochastic process $(X_t)_{t \in T}$ is a Gaussian process if the marginal distribution of $X_{t_1}, \ldots, X_{t_n}$ is a multivariate Gaussian distribution for all $t_1, \ldots, t_n \in T$. Let ...
- 153
1
vote
0
answers
130
views
Cokriging, zero distance semivariance [gstat]
Trying cokriging with simulated data, I faced a problem that did not seem one in the demo(cokriging) with the meuse dataset:
I can't use ...
0
votes
1
answer
674
views
Correct way to compute the error variance of ordinary kriging
I'm learning ordinary kriging and I found some discrepancies in the method of computing error variance among different materials I read, in general there are 2 different formula:
Method 1:
...
- 103
1
vote
1
answer
247
views
spacetime R: How to handle missing data in a space-time-full data structure for spatio-temporal kriging purposes?
I am using R and the spacetime package.
I am having problems using STFDF.
I want to use the STF data-structure since I have spacetime data with recurrent observations for fixed spatial coordinates. ...
- 163
2
votes
0
answers
31
views
Noisy conditional simulation
A conditional random field $Z_C(x)$ is a random field whose realisations $z_C(x)$ always take the same values $z_C(x_a)$ at locations $x_a$. Realisations of $Z_C(x)$ can be produced as follows (...
- 1,145
10
votes
1
answer
6k
views
Ordinary kriging example step by step?
I have followed tutorials online for spatial kriging with both geoR and gstat (and also ...
- 163
7
votes
2
answers
9k
views
What is the nugget effect?
I don't understand exactly what is meant by the term "nugget effect" in geostatistics. When looking at empirical variograms plotting the variogram $\gamma(h)$ vs. the lag $h$, the nugget is defined as ...
- 2,806
1
vote
1
answer
91
views
Are Kriging's residuals (i.e. $Z-\hat{Z}$) spatially independent?
Suppose data ${Z(s_i ):i=1, ..., n}$ are observed at spatial locations ${s_i :i=1, ..., n}$. To carry out the spatial prediction (predict un unknown $Z(s_0)$ at a known location $s_0$) we can use a ...
- 31
2
votes
1
answer
1k
views
Is a function describable by a Gaussian process smooth?
I understand that a stochastic process or function is considered a Gaussian process if sampling from it at any point some set of times yields a set of observations that match a Gaussian random ...
- 1,197
0
votes
1
answer
283
views
How do you interpret this variogram?
Description:
8000 spatial data points spanned over an entire state
200 bins are used
My question: Is the variogram telling something about the nature of the data? Why is it fluctuating? Should I do ...
-1
votes
1
answer
40
views
What is the statistically correct way of performing the following interpolation?
I have a table with thousands of entries similar to the following:
It is desired to determine Property 2 at the unknown locations. Property 1 and Property 2 are spatially correlated. By that I mean ...
1
vote
0
answers
733
views
Spherical vs. Exponential Kriging Covariance Functions
A statistical epidemiologist colleague of mine told me that in comparing spherical vs exponential kriging covariance functions, only the latter (i.e., exponential) function is generally a valid model.
...
- 1,413