plot

Plots on matplotlib (the plot extra).

Every function draws on ax when given, else on a new figure, and returns (fig, ax); scatter_matrix and variograms take and return a grid of axes instead, and category_colors and category_legend return what they make.

InvalidInput

histogram

histogram(
    values,
    *,
    weights=None,
    bins=40,
    log=False,
    stats=False,
    data=None,
    ax=None,
    **kwargs,
)

Histogram of relative frequencies, declustered by weights; log bins on a log axis.

Parameters:

Name	Type	Description	Default
values	str or array_like	Values; NaN is ignored.	required
weights	str or array_like	Declustering weights.	None
bins	int or array_like	Number of bins or their edges.	40
log	bool	Log-spaced bins over the positive values and a log x axis.	False
stats	bool	Write the count, weighted mean, CV and P10, P50, P90 in the upper right corner.	False
data	(PointSet, BlockModel, Table or mapping)	Container whose columns values and weights may name; a block model's volumes are the default weights.	None
**kwargs		Passed to ax.hist.	required

probability

probability(
    values,
    *,
    weights=None,
    log=False,
    cap=None,
    fences=None,
    data=None,
    ax=None,
    **kwargs,
)

Cumulative probability on a normal scale: a Gaussian (or, with log, lognormal) distribution is a line.

Parameters:

Name	Type	Description	Default
values	str or array_like	Values; NaN is ignored.	required
weights	str or array_like	Declustering weights.	None
log	bool	Log x axis.	False
cap	float	Top cut, drawn as a dashed vertical line.	None
fences	float	Outlier fences P25 - fences × IQR and P75 + fences × IQR (1.5 is customary, 3 for far outliers), from weighted quartiles of the values, or of their logarithm with log; drawn as dotted vertical lines labeled with the count of values beyond each.	None
data	(PointSet, BlockModel, Table or mapping)	As in histogram.	None
**kwargs		Passed to ax.plot.	required

cdf

cdf(
    values,
    *,
    weights=None,
    labels=None,
    log=False,
    stats=False,
    data=None,
    ax=None,
    **kwargs,
)

Cumulative distribution of one or several series overlaid, e.g. domains, or clustered and declustered.

Parameters:

Name	Type	Description	Default
values	str, array_like or list of them	One series, or several; NaN is ignored.	required
weights	array_like or list	Declustering weights: an array for one series; for several, one array or None each.	None
labels	list of str	Legend entries.	None
log	bool	Log x axis.	False
stats	bool	Write the count, weighted mean, CV and P10, P50, P90 of each series in the upper left corner, which a cumulative distribution leaves empty.	False
data	(PointSet, BlockModel, Table or mapping)	Container whose columns values and weights may name; a block model's volumes are the default weights.	None
**kwargs		Passed to every ax.step.	required

qq

qq(
    x,
    y,
    *,
    x_weights=None,
    y_weights=None,
    quantiles=None,
    log=False,
    data=None,
    ax=None,
    **kwargs,
)

Quantiles of y against the same quantiles of x, with the 1:1 line on which equal distributions lie.

Parameters:

Name	Type	Description	Default
x	str or array_like	Two samples, e.g. one domain and another, or composites and blocks; NaN is ignored.	required
y	str or array_like	Two samples, e.g. one domain and another, or composites and blocks; NaN is ignored.	required
x_weights	str or array_like	Declustering weights of each.	None
y_weights	str or array_like	Declustering weights of each.	None
quantiles	array_like	Probabilities; default 0.01 to 0.99 in steps of 0.01.	None
log	bool	Log axes.	False
data	(PointSet, BlockModel, Table or mapping)	As in histogram, for both samples.	None
**kwargs		Passed to ax.plot.	required

boxplot

boxplot(
    values,
    categories,
    *,
    weights=None,
    sort=False,
    log=False,
    scheme=None,
    data=None,
    ax=None,
    **kwargs,
)

One box per category: P25 to P75, median, P10 to P90 whiskers and the mean, all weighted.

Parameters:

Name	Type	Description	Default
values	str or array_like	Values; NaN is ignored.	required
categories	array_like	Category (e.g. domain) of each value, its code when scheme is given; each box is labeled with its count.	required
weights	str or array_like	Declustering weights.	None
sort	bool	Order boxes by median, else by category.	False
log	bool	Log value axis.	False
scheme	Categories	Names and colors of the categories.	None
data	(PointSet, BlockModel, Table or mapping)	Container whose columns values, categories and weights may name; a block model's volumes are the default weights.	None
**kwargs		Passed to ax.bxp.	required

variogram

variogram(
    experimental,
    *,
    variogram=None,
    direction=None,
    ax=None,
    **kwargs,
)

Experimental variogram, points sized by pair count, with variogram and its sill.

Parameters:

Name	Type	Description	Default
experimental	ExperimentalVariogram		required
variogram	Variogram		None
direction	tuple of float	(azimuth, dip) in degrees along which an anisotropic variogram is drawn; without it, lags are taken as distances in the model's isotropic space.	None
**kwargs		Passed to ax.scatter and, for its color, to the model line.	required

variograms

variograms(
    variograms,
    *,
    model=None,
    labels=None,
    axes=None,
    **kwargs,
)

Every direct and cross variogram of a VariogramSet as a matrix of panels, with model if given.

Panel (i, j) holds the direct variogram of variable i on the diagonal and the cross variogram of i and j above it; the lower triangle is left empty. Points are sized by their share of the largest pair count, one color per direction.

Parameters:

Name	Type	Description	Default
variograms	VariogramSet		required
model	Coregionalization	Drawn as C_ij(0) - C_ij(h) along each direction, or east when the set is omnidirectional.	None
labels	list of str	Variable names; default the set's column names or indices.	None
axes	array of Axes	(nvar, nvar) axes to draw on; default a new figure.	None
**kwargs		Passed to every ax.scatter.	required

Returns:

Name	Type	Description
fig	Figure
axes	ndarray of Axes	(nvar, nvar).

histogram_reproduction

histogram_reproduction(
    check,
    *,
    variable=0,
    scores=False,
    band=(0.0, 1.0),
    ax=None,
    **kwargs,
)

Cumulative distributions of the realizations as a band, against the declustered data's; for categories, the spread of each category's proportion across realizations against the data's.

Parameters:

Name	Type	Description	Default
check	RealizationCheck	Result of check_realizations.	required
variable	int or str	Variable to draw.	0
scores	bool	In the normal scores of the data, against the standard normal distribution, instead of data units.	False
band	tuple of float	Quantiles across realizations bounding the band; default their full range.	(0.0, 1.0)
**kwargs		Passed to ax.fill_betweenx (ax.vlines for categories).	required

variogram_reproduction

variogram_reproduction(
    check, *, variable=0, band=(0.0, 1.0), ax=None, **kwargs
)

Experimental variograms of the realizations as a band per direction, with the data's as points and the model as a line.

Parameters:

Name	Type	Description	Default
check	RealizationCheck	Result of check_realizations with lag and max_lag.	required
variable	int or str	Variable, or category for its indicator, to draw.	0
band	tuple of float	Quantiles across realizations bounding each band; default their full range.	(0.0, 1.0)
**kwargs		Passed to every ax.fill_between.	required

correlation_reproduction

correlation_reproduction(
    check, *, band=(0.0, 1.0), ax=None, **kwargs
)

Correlation of each pair of variables across realizations as a range, against the declustered data's.

Parameters:

Name	Type	Description	Default
check	RealizationCheck	Result of check_realizations with several variables.	required
band	tuple of float	Quantiles across realizations bounding each range; default their full range.	(0.0, 1.0)
**kwargs		Passed to ax.vlines.	required

variogram_volume

variogram_volume(
    volume,
    *,
    plane="major-semi",
    ellipse=True,
    ax=None,
    **kwargs,
)

Slice of a variogram volume through two of its principal axes, with the range ellipse.

Parameters:

Name	Type	Description	Default
volume	VariogramVolume		required
plane	(major - semi, major - minor, semi - minor)	Principal axes spanning the slice, drawn along x and y.	"major-semi"
ellipse	bool	Draw the range ellipse and the two axes.	True
**kwargs		Passed to ax.pcolormesh (e.g. cmap, vmax).	required

scatter

scatter(x, y, *, line=True, data=None, ax=None, **kwargs)

y against x with the 1:1 line and, if line, the least-squares regression of y on x.

The regression slope of true on estimated values below 1 flags conditional bias. x and y may name columns of data.

ternary

ternary(
    data,
    *,
    parts=None,
    labels=None,
    grid=True,
    ax=None,
    **kwargs,
)

Compositions of three parts as points in an equilateral triangle, each vertex a pure part.

Parameters:

Name	Type	Description	Default
data	(array_like, mapping, Table, PointSet or BlockModel)	(n, 3) parts, or a container holding the parts columns. Each row is closed before plotting, so the parts may be any three of a larger composition (a subcomposition).	required
parts	list of 3 str	Columns of data, in vertex order: bottom left, bottom right, top.	None
labels	list of 3 str	Vertex labels; default the part names.	None
grid	bool	Draw lines at every 20 % of each part.	True
**kwargs		Passed to ax.scatter.	required

biplot

biplot(data, *, parts=None, labels=None, ax=None, **kwargs)

Covariance biplot of the clr coordinates: samples as points, parts as rays from the origin.

Ray length is the standard deviation of a part's clr coordinate; the distance between two ray tips is the standard deviation of the log-ratio of the two parts, so tips close together mark parts in near-constant proportion. The axes are the first two principal components, labeled with the share of total variance they hold.

Parameters:

Name	Type	Description	Default
data	(array_like, mapping, Table, PointSet or BlockModel)	(n, D) positive parts, or a container holding the parts columns; rows with a missing part are skipped.	required
parts	list of str	Columns of data; default all of them.	None
labels	list of str	Ray labels; default the part names.	None
**kwargs		Passed to ax.scatter.	required

correlation

correlation(
    data,
    *,
    columns=None,
    labels=None,
    weights=None,
    method="pearson",
    colorbar=True,
    ax=None,
    **kwargs,
)

Correlation, rank correlation or covariance matrix as a heatmap, each cell written out.

Parameters:

Name	Type	Description	Default
data	(array_like, mapping, Table, PointSet or BlockModel)	(n, d) values, or named columns; each pair uses the rows where both are present, as boitata.correlation does.	required
columns	list of str	Columns of data to draw, in order; default all of them.	None
labels	list of str	Axis labels; default the column names or indices.	None
weights	array_like or str	Declustering weights, or their column in data.	None
method	(pearson, spearman, covariance)	What to show; correlations span -1 to 1, covariances ± their largest magnitude.	"pearson"
colorbar	bool	Add a color bar.	True
**kwargs		Passed to ax.imshow (e.g. cmap).	required

declustering

declustering(result, *, naive=None, ax=None, **kwargs)

Declustered mean against cell size, the chosen size marked, and the naive mean for reference.

Parameters:

Name	Type	Description	Default
result	Declustering	Result of boitata.cell_declustering scanning cell sizes.	required
naive	float	Mean of the values without weights, drawn as a dotted line.	None
**kwargs		Passed to ax.plot.	required

conditional

conditional(
    x,
    y,
    *,
    bins=10,
    weights=None,
    log=False,
    data=None,
    ax=None,
    **kwargs,
)

Scatter of y against x with the mean and P10 to P90 of y in bins of x, drawn at each bin's median.

Parameters:

Name	Type	Description	Default
x	str or array_like	Pairs of values; pairs with NaN are ignored.	required
y	str or array_like	Pairs of values; pairs with NaN are ignored.	required
bins	int or array_like	Number of bins holding about as many pairs each, or bin edges of x.	10
weights	str or array_like	Declustering weights.	None
log	bool	Log axes; non-positive values are then ignored.	False
data	(PointSet, BlockModel, Table or mapping)	Container whose columns x, y and weights may name; a block model's volumes are the default weights.	None
**kwargs		Passed to ax.scatter.	required

completeness

completeness(data, *, columns=None, ax=None, **kwargs)

Rows by number of variables present, the complete rows in the accent color, each bar labeled.

Parameters:

Name	Type	Description	Default
data	(array_like, mapping, Table, PointSet or BlockModel)	(n, d) values, or named columns; NaN is missing.	required
columns	list of str	Columns of data to draw, in order; default all of them.	None
**kwargs		Passed to ax.bar.	required

scatter_matrix

scatter_matrix(
    data,
    *,
    columns=None,
    labels=None,
    weights=None,
    log=False,
    bins=30,
    axes=None,
    **kwargs,
)

Pairwise scatters of the columns of data, with their histograms on the diagonal.

Each scatter is annotated with the Pearson (r) and rank correlation of its pair, weighted by weights and over the rows where both values are present, as boitata.correlation computes them.

Parameters:

Name	Type	Description	Default
data	(array_like, mapping, Table, PointSet or BlockModel)	(n, d) values, or named columns; NaN is ignored pair by pair.	required
columns	list of str	Columns of data to draw, in order; default all of them.	None
labels	list of str	Axis labels; default the column names or indices.	None
weights	array_like or str	Declustering weights, or their column in data, used by the histograms and the correlations.	None
log	bool or sequence of bool	Log axes and bins, for every column or per column; non-positive values are then ignored.	False
bins	int	Number of histogram bins.	30
axes	array of Axes	(d, d) axes to draw on; default a new figure.	None
**kwargs		Passed to every ax.scatter.	required

Returns:

Name	Type	Description
fig	Figure
axes	ndarray of Axes	(d, d); axes[i, j] has column j across and column i up.

section

section(
    model,
    values,
    *,
    axis="z",
    index=None,
    plane=None,
    origin=None,
    azimuth=90.0,
    dip=90.0,
    holes=None,
    width=None,
    resolution=None,
    colorbar=True,
    scheme=None,
    ax=None,
    **kwargs,
)

Slice of a block model across axis at cell index (default: the middle), or on any plane, at true scale, with the drill holes near it projected on top.

Parameters:

Name	Type	Description	Default
model	BlockModel	Missing blocks are left blank. Drawn as true block edges (exact for masked and sub-blocked layouts) whenever the cut is normal to one of the model's own local axes: always along axis, and for an explicit plane only when it matches the model's rotation. Otherwise resampled onto a raster.	required
values	str or array_like	Column name, or one value per block.	required
axis	(x, y, z)	Axis normal to the slice.	"x"
index	int	Cell index along axis.	None
plane	tuple	(center, azimuth, dip) as in slab; replaces axis and index. The model is sampled on a raster in section coordinates.	None
origin	array_like	(x, y, z) point on the plane; with azimuth and dip, the same as plane=(origin, azimuth, dip).	None
azimuth	float	Bearing of the section line and dip of the plane through origin, in degrees (90 and 90: a vertical east-west section).	90.0
dip	float	Bearing of the section line and dip of the plane through origin, in degrees (90 and 90: a vertical east-west section).	90.0
holes	Drillholes	Traces within width of the plane, clipped to that slab, projected and labeled at their shallowest point in it.	None
width	float	Full width of the slab around the plane that keeps holes; default the largest block edge.	None
resolution	float	Raster step on plane; default half the smallest block edge.	None
colorbar	bool	Add a color bar labeled with the column name, or a legend with scheme.	True
scheme	Categories	values are codes of these categories, drawn in their colors.	None
**kwargs		Passed to ax.imshow (e.g. cmap, norm, vmin).	required

fence

fence(
    model,
    values,
    positions,
    *,
    azimuth,
    dip=90.0,
    resolution=None,
    colorbar=True,
    scheme=None,
    axes=None,
    **kwargs,
)

One vertical (or dipping) section per row of positions, on the same azimuth and dip, sharing one color scale: a fence of parallel sections stepped along a corridor.

Parameters:

Name	Type	Description	Default
model	BlockModel	As in section.	required
values	str or array_like	Column name, or one value per block.	required
positions	array_like	(n, 3) plane centers, one per panel, as section's plane center.	required
azimuth	float	Orientation shared by every panel's plane, in degrees; dip as in section's plane (90 vertical).	required
dip	float	Orientation shared by every panel's plane, in degrees; dip as in section's plane (90 vertical).	required
resolution	float	Raster step on each plane; default half the smallest block edge.	None
colorbar	bool	Add one color bar for every panel, labeled with the column name, or a legend with scheme.	True
scheme	Categories	values are codes of these categories, drawn in their colors.	None
axes	array of Axes	One per position; default a new figure.	None
**kwargs		Passed to every ax.imshow (e.g. cmap, norm, vmin, vmax).	required

Returns:

Name	Type	Description
fig	Figure
axes	ndarray of Axes

slab

slab(
    coords,
    values=None,
    *,
    plane,
    thickness,
    meshes=None,
    lines=None,
    labels=None,
    colorbar=True,
    scheme=None,
    ax=None,
    **kwargs,
)

Points within a slab around a plane, in section coordinates, with mesh traces, clipped lines and labels.

Section coordinates are world coordinates along the strike and up the dip of the plane, so a vertical east-west section reads easting and elevation, and a horizontal plane easting and northing.

Parameters:

Name	Type	Description	Default
coords	PointSet or array_like	(n, 3) coordinates.	required
values	str or array_like	Column of coords, or one value per point, coloring them; default one color.	None
plane	tuple	(center, azimuth, dip): a point on the plane, the bearing of the section line and the dip of the plane (90 is vertical, 0 a plan), in degrees.	required
thickness	float	Full width of the slab.	required
meshes	Mesh or list of Mesh	Drawn as their intersection with the plane.	None
lines	list of array_like	(m, 3) polylines, e.g. drill-hole traces, clipped to the slab.	None
labels	sequence of str	One per point; each distinct label is written once, above its highest point in the slab.	None
colorbar	bool	With values, add a color bar labeled with the column name, or a legend with scheme.	True
scheme	Categories	values are codes of these categories, drawn in their colors.	None
**kwargs		Passed to ax.scatter (e.g. s, cmap, norm, color).	required

swath

swath(swaths, *, labels=None, y='mean', ax=None, **kwargs)

Mean, tonnage or metal per slice of one or several swath results, with the first one's counts as light bars.

Parameters:

Name	Type	Description	Default
swaths	Table or list of Table	Results of boitata.swath, e.g. composites and blocks with the same width.	required
labels	list of str	Legend entries.	None
y	(mean, tonnage, metal)	What to draw per slice.	"mean"
**kwargs		Passed to every ax.plot.	required

category_colors

category_colors(scheme)

Color map and norm drawing code i of scheme in its color i.

Parameters:

Name	Type	Description	Default
scheme	Categories	Colors of the categories; when it has none, spread over matplotlib's image.cmap with other in gray.	required

Returns:

Name	Type	Description
cmap	ListedColormap
norm	BoundaryNorm	Boundaries at every code ± 0.5; pass both as cmap= and norm= to imshow or scatter.

category_legend

category_legend(scheme, ax, **kwargs)

Legend of scheme: one patch per category, in code order, in the colors of category_colors.

Parameters:

Name	Type	Description	Default
scheme	Categories		required
ax	Axes or Figure	Where the legend goes.	required
**kwargs		Passed to ax.legend (e.g. loc, ncol).	required

Returns:

Type	Description
Legend

category_swath

category_swath(
    coords,
    categories,
    width,
    *,
    azimuth=None,
    axis=None,
    weights=None,
    colors=None,
    scheme=None,
    ax=None,
    **kwargs,
)

Proportion of each category per slice along a direction, as stacked bars.

Parameters:

Name	Type	Description	Default
coords	(PointSet, BlockModel or array_like)	(n, 2) or (n, 3) coordinates; a container's columns may be named by categories and weights.	required
categories	str or array_like	Category (e.g. lithology) of each sample; its code, NaN for none, when scheme is given.	required
width		Slices as in boitata.swath: width along azimuth (degrees from north) or axis ("x", "y", "z").	required
azimuth		Slices as in boitata.swath: width along azimuth (degrees from north) or axis ("x", "y", "z").	required
axis		Slices as in boitata.swath: width along azimuth (degrees from north) or axis ("x", "y", "z").	required
weights	str or array_like	Declustering weights or lengths.	None
colors	sequence	One color per category, sorted; default the scheme's, else spread over matplotlib's image.cmap with other in gray.	None
scheme	Categories	Order, names and colors of the categories.	None
**kwargs		Passed to every ax.bar.	required

proportions

proportions(
    categories,
    *,
    weights=None,
    scheme=None,
    data=None,
    ax=None,
    **kwargs,
)

Proportion of each category as horizontal bars, weighted, with the unweighted proportions as ticks.

Parameters:

Name	Type	Description	Default
categories	array_like	Category of each sample; its code, NaN for none, when scheme is given.	required
weights	str or array_like	Declustering weights or lengths; the bars are then weighted and a dark tick marks each unweighted share.	None
scheme	Categories	Order, names and colors of the categories.	None
data	(PointSet, BlockModel, Table or mapping)	Container whose columns categories and weights may name; a block model's volumes are the default weights.	None
**kwargs		Passed to ax.barh.	required

strip_log

strip_log(
    drillholes,
    hole,
    *,
    columns=(),
    categories=(),
    scheme=None,
    runs=None,
    ax=None,
    **kwargs,
)

Holes as vertical logs side by side: category bars, grade profiles as steps and ore runs shaded.

Parameters:

Name	Type	Description	Default
drillholes	Drillholes	With an interval table.	required
hole	str or sequence of str	Hole or holes to draw, left to right.	required
columns	sequence of str	Grade columns, one track each, drawn as steps from zero to the largest value over the holes drawn.	()
categories	str or sequence of str	Categorical columns (e.g. lithology), one bar each.	()
scheme	Categories or dict	Names and colors of the categories, or a Categories per column name; values outside it are left blank. Without one, values are sorted and spread over matplotlib's image.cmap.	None
runs	Table	From Drillholes.runs; ore runs are shaded across the grade tracks.	None
**kwargs		Passed to every grade ax.plot.	required

holes

holes(
    paths,
    *,
    plane=None,
    mark=None,
    hole="HOLE_ID",
    labels=True,
    ax=None,
    **kwargs,
)

Drill-hole traces, one line per hole, in plan or projected on a section.

Parameters:

Name	Type	Description	Default
paths	Table	Result of Drillholes.paths: a hole id column, depth, x, y and z.	required
plane	tuple	(center, azimuth, dip) as in section's plane; projects traces into that section instead of drawing plan-view (x, y).	None
mark	array_like	(m, 3) extra points scattered on top, e.g. flagged stations from Drillholes.at.	None
hole	str	Column of paths naming each row's hole.	'HOLE_ID'
labels	bool	Label each hole at its shallowest row.	True
**kwargs		Passed to every ax.plot.	required

directions

directions(
    anisotropy,
    *,
    plane=None,
    thickness=None,
    ax=None,
    **kwargs,
)

Major axis of each local anisotropy as a line through its location, on a plan or a section.

Each line is the major axis projected on the view, so an axis plunging out of it draws short.

Parameters:

Name	Type	Description	Default
anisotropy	LocalAnisotropy	Locations and angles (azimuth, dip, rake).	required
plane	tuple	(center, azimuth, dip) of a section as in slab; default a plan.	None
thickness	float	With plane, full width of the slab of locations drawn; default all.	None
**kwargs		Passed to ax.quiver (e.g. scale, width, color).	required

paired_bias

paired_bias(bias, *, ax=None, **kwargs)

Relative bias of b over a per bin of pairing distance, with the pair counts as light bars.

Parameters:

Name	Type	Description	Default
bias	Table	Result of boitata.paired_bias.	required
**kwargs		Passed to ax.plot.	required

uncertainty_curve

uncertainty_curve(
    curve, *, threshold=0.15, required=None, ax=None
)

Uncertainty quantiles against data spacing, the share within threshold on the right axis.

Parameters:

Name	Type	Description	Default
curve	Table	Result of boitata.uncertainty_curve; its P... columns are drawn, the first solid, the rest dashed.	required
threshold	float	Acceptable uncertainty, drawn as a horizontal line.	0.15
required	float	Required spacing, e.g. from boitata.required_spacing, marked on the threshold line.	None

grade_tonnage

grade_tonnage(table, *, relative=False, ax=None, **kwargs)

Tonnage (solid) and mean grade (dashed, right axis) above cutoff, for one curve or several in one color each.

Parameters:

Name	Type	Description	Default
table	Table or dict of str to Table	Result of boitata.grade_tonnage, HermiteAnamorphosis.grade_tonnage, UniformConditioning.grade_tonnage, boitata.compare_models or SimulationSummary.grade_tonnage; rows are split into curves by their model and category columns. A dict names several such tables. With a probability column, each curve is drawn at the probability nearest 0.5, in a band between the lowest and highest.	required
relative	bool	Tonnage as a fraction of each curve's tonnage at its lowest cutoff, to compare samples with blocks.	False
**kwargs		Passed to every ax.plot.	required

cross_validation

cross_validation(
    cv, *, kind="scatter", coords=None, ax=None, **kwargs
)

Cross-validation as actual against estimate, as an accuracy plot, or as its errors.

Parameters:

Name	Type	Description	Default
cv	CrossValidation or IndicatorCrossValidation	Result of cross_validate.	required
kind	(scatter, accuracy, errors)	"scatter": actual against estimate with the 1:1 line, the regression of actual on estimate (a slope below 1 is conditional bias) and the statistics. "accuracy": fraction of samples inside their symmetric p-probability interval against p, with the goodness statistic; on or above the diagonal is accurate. Kriging takes the Gaussian interval of its estimate and variance, indicator kriging its corrected distribution. "errors": estimate minus actual mapped at coords, or without them a histogram.	"scatter"
coords	PointSet or array_like	Sample locations for the error map, in the order given to fit.	None
**kwargs		Passed to ax.scatter (scatter, error map), ax.plot (accuracy) or ax.hist (histogram).	required

contact

contact(
    table,
    *,
    labels=("inside", "outside"),
    ax=None,
    **kwargs,
)

Mean grade against signed distance to a contact, the sample counts as light bars; a jump at zero is a hard contact, a gradual change a soft one.

Parameters:

Name	Type	Description	Default
table	Table	Result of boitata.contact; negative distances are inside.	required
labels	tuple of str	Names of the inside and outside domains, written on either side.	('inside', 'outside')
**kwargs		Passed to ax.plot.	required

domain_change

domain_change(
    table,
    *,
    value="tonnage",
    relative=False,
    fmt=None,
    ax=None,
    **kwargs,
)

Cross-table of two categorical models as a matrix, rows the first model's classes, columns the second's, each cell labeled with its value. The diagonal, what is unchanged, is outlined and left blank so the grays scale to what moves.

Parameters:

Name	Type	Description	Default
table	Table	Result of boitata.domain_change.	required
value	str	Column to show: "tonnage", "metal" or "mean_grade".	'tonnage'
relative	bool	Each row as a fraction of its total: the share of each class of the first model going to each class of the second (not with "mean_grade").	False
fmt	str	Format of the cell labels; default "{:.0%}" when relative, else "{:,.3g}".	None
**kwargs		Passed to ax.imshow.	required

transition_mds

transition_mds(table, *, ax=None, **kwargs)

Classical (Torgerson) MDS of an along-hole transition table: categories that transition into each other most often sit close together in 2D.

The frequency column, one class's row shares, is symmetrized into a dissimilarity, 1 - (freq + freq.T) / 2, then scaled from the top two eigenvectors of its double-centered squared distances.

Parameters:

Name	Type	Description	Default
table	Table	Result of boitata.transition_matrix.	required
**kwargs		Passed to ax.scatter.	required

uncertain

uncertain(
    values,
    uncertainty,
    *,
    model=None,
    axis="z",
    index=None,
    plane=None,
    resolution=None,
    extent=None,
    cmap=None,
    norm=None,
    scheme=None,
    label=None,
    legend_ax=None,
    ax=None,
)

Image whose color gives a value and whose fading towards white gives how uncertain it is.

The legend is a fan: the value runs across its angle, certainty along its radius, from white at the center (uncertain) to the full color on the arc (certain).

Parameters:

Name	Type	Description	Default
values	str or array_like	2D image, e.g. the mean of the realizations on a section; NaN is left blank. With model, a column name or one value per block.	required
uncertainty	str or array_like	Same shape as values, from 0 (certain) to 1 (no information), e.g. the realizations' standard deviation over the global one. Clipped to [0, 1].	required
model	BlockModel	Slice it as section does; missing blocks are left blank.	None
axis	(x, y, z)	With model, axis normal to the slice.	"x"
index	int	With model, cell index along axis (default: the middle).	None
plane	tuple	With model, (center, azimuth, dip) as in slab, replacing axis and index.	None
resolution	float	Raster step on plane; default half the smallest block edge.	None
extent	tuple of float	Passed to ax.imshow; set from model when given.	None
cmap	str or Colormap	Default: matplotlib's image.cmap.	None
norm	Normalize	Maps values to [0, 1]; default spans their range.	None
scheme	Categories	values are codes of these categories (e.g. the most likely one), drawn in their colors and keyed by a legend instead of the fan; replaces cmap and norm.	None
label	str	Value name written under the legend, or its title with scheme.	None
legend_ax	Axes	Where to draw the legend; default below ax.	None

describe builtin

describe(
    values, *, weights=None, quantiles=Ellipsis, data=None
)

Weighted (declustered) statistics of values; NaN is skipped.

Parameters:

Name	Type	Description	Default
values	array_like or str	Values, or their column in data.	required
weights	array_like or str	Declustering weights; default the volumes of a BlockModel data, else 1.	None
quantiles	sequence of float	Probabilities of the quantile keys.	Ellipsis
data	(PointSet, BlockModel, Table or dict)	Where column names are looked up.	None

Returns:

Type	Description
dict	n, mean, variance, std, cv, min, max and one key per quantile named P10, P97.5, ... (mid-point cumulative weights).

normal_cdf builtin

normal_cdf(x)

Standard normal CDF.

normal_ppf builtin

normal_ppf(p)

Standard normal quantile function.

used in

• From drill holes to a 3D model
• Describing data
• Spatial continuity
• Simulation
• Checking a model
• Flag solid proportions in a block model
• Sub-block a model from solids
• Compare two estimates of one deposit