Skip to content

which page do i need?

find your question and follow the link. learn pages explain an idea from scratch, examples show one task on real data, workflows follow a whole problem from question to decision, and API pages list every argument.

new to geostatistics

question read
what does a sample represent, and why does its size matter? learn 1. samples and support
how do i summarize a variable when the drilling is clustered? learn 2. describing data
what is a variogram and how do i read one? learn 3. spatial continuity
how does kriging choose its weights? learn 4. kriging
why simulate when i can krige? learn 5. simulation
how do i know a model is any good? learn 6. checking a model
how is the library laid out? guide how boitatá is organized
i want one pass from samples to a checked estimate. quick tour

drill holes, solids and grids

question read
i have a new drill hole database. is it clean? workflow audit drill holes before modeling, then checking drill holes
my holes have overlapping or missing intervals. checking drill holes (check_drillholes, fix_drillholes)
where is each sample in 3D? desurveying drill holes
my assays have different lengths. compositing
some collars or samples appear twice. duplicates
an interval straddles a geological contact. mesh-crossing interval splits
where does each hole run above cutoff? runs and strip logs
i need a block model that covers my data. block models, block model from extents
which blocks fall inside a solid, and by how much? solids, workflow flag solid proportions
a thin solid gets the wrong volume on whole blocks. sub-blocks, and the sub-blocking tutorial in workflows
my solid has holes, flipped faces or zero volume. workflow find and fix degenerate solids
my rock-type model is full of isolated specks. domain cleanup
i have a lease boundary or pit outline. polygons
i need to read or write OBJ, STL, DXF, shapefiles or GeoTIFF. mesh files, GIS formats
my block model does not fit in memory. models larger than memory
i want to see holes and blocks in 3D or on sections. 3D views, stepped sections, hole traces, sections with holes, live scene

looking at the data

question read
the rich zones were drilled more densely. what is the true mean? declustering, weight declustering
how do the statistics differ between domains? statistics by domain
a few extreme assays dominate the mean. top cuts, capping transform
many samples share one value, such as the detection limit. despiking, censored normal score
is the boundary between two domains sharp or gradual? contacts, soft-boundary statistics
do two sampling campaigns agree? paired data
does the grade drift along strike or with depth? swaths
how closely spaced is my drilling? data spacing
how sure am i of the mean itself? spatial bootstrap
how do my variables relate to each other? correlations

transforms

question read
i need gaussian scores for simulation or multigaussian kriging. normal-score transform
my grades are parts of a whole and sum to 100 %. compositional data
i have several correlated grades to decorrelate. multivariate transforms
some samples lack one of the assays. imputation, spatial imputation
my layer undulates between two surfaces. unfolding
question read
how do i compute a variogram along a direction? experimental variograms
how do i fit a model to it? variogram fitting
how large is the nugget effect? downhole nugget, nugget inference
i have two or more variables to model together. coregionalization
the direction of continuity changes across the deposit. local variogram parameters, locally varying anisotropy
how big should my search be? from variogram to search plan, search, search calibration

estimating

question read
i want a first estimate of a grade. ordinary kriging
i want inverse distance or nearest neighbor to compare against. simple estimators
i need block averages, not point values. block kriging
the grade has a trend. universal kriging, external drift kriging
high grades spread too far. high-grade restriction
a secondary variable is sampled more densely. cokriging
i want the probability of exceeding a cutoff. indicator kriging, multiple indicator kriging, multigaussian kriging
samples across a domain boundary should count a little. soft boundaries

categories and domains

question read
how do i name and color rock codes? categories
how much did my domains change between two models? domain change tables
what proportion of each rock type is in each block? local proportions, categorical indicator kriging
i need possible layouts of the rock types. sequential indicator simulation, plurigaussian simulation, grades in simulated rock types

simulating and recoverable resources

question read
i want many equally likely grade maps. sequential gaussian simulation, turning bands, direct sequential simulation
my grid is large and simulation is slow. SGS along a shared path
i need simulated values at block size. simulation at block support
several grades must stay correlated. multivariate simulation, collocated cosimulation
geology looks like channels or lenses that a variogram cannot draw. SNESIM and its multigrid, training images, image quilting
how much ore is above cutoff at the size of a mining block? discrete gaussian model, uniform conditioning
i need a grade-tonnage curve and the metal it implies. workflow grade-tonnage and metal change, result plots

checking a model

question read
does my estimate honor the data? model checks, section validation plates
are my variogram and search any good? cross-validation
which blocks are well informed? kriging diagnostics
do my realizations reproduce the histogram and variogram? realization checks
how do i classify the resource? classification
i have two estimates of one deposit. which differs where? workflow compare two estimates

surfaces, whole studies and files

question read
i want a grade shell or a contact surface from drill holes. grade shells, contact surfaces, layered surfaces
i have measured orientations of the contact. structural data
i want to see a full study from raw tables to a result. case studies, such as from drill holes to a classified model
i want to save a variogram, a transform or a fitted estimator. guide saving models, saving and loading, parquet
what does an argument do? API: containers, drill holes, EDA, transforms, variography, estimation, simulation, modeling, datasets and plots

see also

  • workflows: real problems from question to decision.
  • glossary: the terms these pages use.