Note
Click here to download the full example code
along-hole transition matrix and MDS¶
transition_matrix counts which rock type follows which down a hole. for each composite it tallies the class of the
composites one lag deeper in the same hole, into a from/to table shaped like domain_change's: rows the shallower
rock, columns the deeper one, so bt.plot.domain_change draws it unmodified. bt.plot.transition_mds turns the row
frequencies into a dissimilarity and lays the rocks out in 2D by classical MDS, so rocks that often touch sit close
together.
import sys
from pathlib import Path
HERE = Path(__file__).parent if "__file__" in globals() else Path.cwd()
sys.path.insert(0, str(HERE.parents[1]))
import boitata as bt
import numpy as np
from common import save
composites and their rock type¶
5 m composites of the stacked sulphide lenses, majority lithology only (as in domain cleanup); the scheme groups the massive, semi-massive and stringer sulphides into one class.
data = bt.datasets.stacked_sulphide_lenses()
intervals = bt.merge_intervals(data["assays"], data["lithology"])
drillholes = bt.Drillholes(data["collars"], data["surveys"], intervals)
composites = drillholes.composite(5.0, [], categories=["LITH"])
scheme = bt.Categories(
["OB", "HWS", "VCL", "SUL", "FWV"],
mapping={"MS": "SUL", "SMS": "SUL", "STR": "SUL"},
other="DYK",
colors=["#d9d9d9", "#a9bfd3", "#8c8c8c", "#c05a28", "#1f4e79", "#e0c080"],
)
depth = (composites["from"] + composites["to"]) / 2
table = bt.transition_matrix(depth, composites["LITH"], composites["HOLE_ID"], lag=5.0, scheme=scheme)
print(f"{len(composites):,} composites, {int(np.sum(table['count'])):,} pairs 5 m apart")
Out:
23,647 composites, 23,069 pairs 5 m apart
the matrix, reused from domain_change¶
frequency holds each rock's row as shares, so the matrix plot needs no relative=. the diagonal (staying in the
same rock) is outlined and left blank.
fig, ax = bt.plot.domain_change(table, value="frequency", fmt="{:.0%}")
ax.set(xlabel="Rock 5 m deeper", ylabel="Rock at a composite")
save(fig, "matrix")
MDS: who sits next to whom¶
classical MDS turns 1 - (freq + freq.T) / 2 into a 2D layout. the volcaniclastic sits next to the sulphides it
overlies, overburden and hanging wall cluster on their own, and the footwall and dyke sit apart, since either
rarely borders the others.
fig, ax = bt.plot.transition_mds(table)
save(fig, "mds")
Total running time of the script: ( 0 minutes 0.189 seconds)
Download Python source code: example_07_03.py