Skip to content

Commit

Permalink
WIP Variogram definitions
Browse files Browse the repository at this point in the history
  • Loading branch information
TCallaghan2 committed Jun 5, 2024
1 parent ecda191 commit a3887dc
Showing 1 changed file with 33 additions and 5 deletions.
38 changes: 33 additions & 5 deletions PythonScripts/GUI/GeoSAM/InterpolationFrame.py
Original file line number Diff line number Diff line change
Expand Up @@ -367,7 +367,35 @@ def pop_up(self):
This frame allows the user to customize how the interpolation is performed.
Parameters
Variogram Form: This describes the shape of the variogram
Variogram Form:
This defines the shape of the variogram models. The kriging refernces
identify a typical variogram shape. It is a positive sloped function with a
y intercept defined as nugget. The assymptote is defined as the sill. The
inflection point at the sill is defined as the range. Four shapes are
implemented for UK interpolation
• spherical
• exponential
• gaussian
• matern
spherical:
γ(h) = nugget + sill * (3h/2 * range) - 0.5 * (h/range)^3 0 < h <= range
= nugget + sill h > range
= 0 h = 0
exponential:
γ(h) = nugget + sill * (1 - exp(-h/range)) h > 0
= 0 , & h = 0
gaussian
γ(h) = nugget + sill * (1 - exp(-(h/range)^2)) h > 0
= 0 h = 0
matern
γ(h) = nugget + sill
* (1 - {sqrt(2)/Γ(0.5)} * Jn(2, h/range) * sqrt(h/range)) h > 0
= 0 h = 0
where Jn is the Bessel function of the first kind
Spatial Fcn Config File:
This is the name of the file used to hold the spatial function definitions.
Expand All @@ -384,16 +412,16 @@ def pop_up(self):
Spatial Functions
Define non linear spatial functions and paramater search range.
- "Function 1, dim=z, shape=Logistic, precon=0 "
- "Function 2, dim=z, shape=Gaussian, precon=0 "
- "Function 3, dim=x, shape=Logistic, precon=1 "
- Function 1, dim=z, shape=Logistic, precon=0
- Function 2, dim=z, shape=Gaussian, precon=0
- Function 3, dim=x, shape=Logistic, precon=1
These define spatial functions for setting the spatial trend in the
universal kriging algorithm.
The precon=0 term means that the function is not multiplied by another
function. For example,
"Function 3, dim=x, shape=Logistic, precon=1 "
Function 3, dim=x, shape=Logistic, precon=1
indicates that the third function is multiplied by the first function.
This is true for fitting the nonlinear parameters of function 3 hence
the parameters of function 1 must be fit before the parameters of
Expand Down

0 comments on commit a3887dc

Please sign in to comment.