Notebook

Striplog from image¶

We're going to try to make a striplog from an image of a striplog. The impage is Figure 1 from Audrey Estublier*, Alexandre Fornel, Étienne Brosse, Pascal Houel, Jean-Claude Lecomte,Jocelyne Delmas and Olivier Vincké (2017). Simulation of a Potential CO2Storage in the West ParisBasin: Site Characterization and Assessment of theLong-Term Hydrodynamical and Geochemical ImpactsInduced by the CO2Injection. Oil & Gas Science and Technology–Rev. IFP Energies nouvelles (2017) 72, 22. DOI:10.2516/ogst/2017021, licensed CC-BY 4.0.

Fingers crossed!

In [1]:

import matplotlib.pyplot as plt
import numpy as np

This notebook works best with Striplog v0.8.8.

In [2]:

import striplog

striplog.__version__

Out[2]:

'0.8.8'

I checked the PDF and unfortunately the artwork is not stored as vector art — it's a raster. So we'll just have to deal with the various bits of decoration and the low resolution.

Here's the whole page as a PNG.

In [3]:

from IPython.display import Image

Image("./data/ogst160144.png", width=400)

Out[3]:

We need to know the start and end depth, so I cropped out just the striplog itself. It goes from 0 m to 2450 m.

In [4]:

Image("./data/ogst160144_striplog-only.png")

Out[4]:

Some of the decoration might cause problems, let's see.

Make a legend¶

Striplogs are made of Intervals.

Intervals (usually) contain Components.

The Legend is the lookup table for the Components we're going to place in the Intervals.

The colours just have to be close to the colours in the image.

In [5]:

Image("./data/ogst160144_legend.png")

Out[5]:

We need 'Components'. They are defined by property dictionaries, which can contain whatever we want.

In [24]:

from striplog import Legend

legend_text = """colour,width,comp lithology,comp properties,comp role
#509ccc,10,limestone,porous,
#d2e1e8,9,limestone,compact,
#bdc6c8,8,dolomite,compact,
#f8e753,6,sandstone,porous,
#ede8a7,5,sandstone,compact,
#787775,3,marl/shale,,source rock
#a4a09f,4,marl/shale,,seal
"""

legend = Legend.from_csv(text=legend_text)
legend

Out[24]:

component

width

hatch

colour

lithology	limestone
properties	porous

10.0

None

#509ccc

lithology	limestone
properties	compact

9.0

None

#d2e1e8

lithology	dolomite
properties	compact

8.0

None

#bdc6c8

lithology	sandstone
properties	porous

6.0

None

#f8e753

lithology	sandstone
properties	compact

5.0

None

#ede8a7

lithology	marl/shale
role	source rock

3.0

None

#787775

lithology	marl/shale
role	seal

4.0

None

#a4a09f

Making a `striplog`¶

In [27]:

import matplotlib.pyplot as plt
from sklearn.cluster import KMeans

img = plt.imread("./data/ogst160144_striplog-only.png")[:, 10, :3]
kmeans = KMeans(n_clusters=12).fit(img)
quanta = kmeans.cluster_centers_[kmeans.labels_]
plt.imsave("./data/ogst160144_striplog-only_quantized.png", quanta.reshape(-1, 1, 3))

In [28]:

fig, axs = plt.subplots(figsize=(15, 1), ncols=12)
for c, ax in zip(kmeans.cluster_centers_, axs):
    ax.imshow(c.reshape(1,1,3))
    ax.set_title(striplog.utils.rgb_to_hex(c))
    ax.axis('off')

In [29]:

from striplog import Striplog

strip = Striplog.from_img("./data/ogst160144_striplog-only_quantized.png",
                          start=0, stop=2450,
                          col_offset=0,     # How far across the image to extract pixels.
                          legend=legend,
                          tolerance=20,     # How forgiving to be on the colours.
                          row_offset=2,      # How far below a boundary to look at the colour.
                          background='#ffffff',  # Ignore this colour
                         )

# Expect warnings for colours which don't match the legend.

/Users/matt/opt/miniconda3/envs/py39/lib/python3.9/site-packages/striplog/striplog.py:892: UserWarning: from_img() is deprecated; please use from_image()
  warnings.warn(w)
/Users/matt/opt/miniconda3/envs/py39/lib/python3.9/site-packages/striplog/legend.py:897: UserWarning: No match found for #c66242 with tolerance of 20. Best match is Marl/shale, source rock, #787775, d=95.53009996854395
  warnings.warn(w)
/Users/matt/opt/miniconda3/envs/py39/lib/python3.9/site-packages/striplog/legend.py:897: UserWarning: No match found for #d0c070 with tolerance of 20. Best match is Sandstone, porous, #f8e753, d=62.94441992742486
  warnings.warn(w)
/Users/matt/opt/miniconda3/envs/py39/lib/python3.9/site-packages/striplog/legend.py:897: UserWarning: No match found for #363837 with tolerance of 20. Best match is _, #000000, d=95.2732911156112
  warnings.warn(w)
/Users/matt/opt/miniconda3/envs/py39/lib/python3.9/site-packages/striplog/legend.py:897: UserWarning: No match found for #5c6566 with tolerance of 20. Best match is Marl/shale, source rock, #787775, d=36.51027252705737
  warnings.warn(w)

In [30]:

strip

Out[30]:

Striplog(207 Intervals, start=1.6166281755196306, stop=2450.0)

In [31]:

strip = strip.merge_neighbours().prune(limit=2).anneal().merge_neighbours()
strip

Out[31]:

Striplog(143 Intervals, start=1.6166281755196306, stop=2450.0)

In [32]:

strip.plot(legend=legend, lw=1, aspect=20)

Write CSV or LAS3¶

In [11]:

print(strip.to_csv())

Top,Base,Component
7.7596266044340725,32.67211201866978,"Limestone, porous"
32.67211201866978,74.32905484247375,"Sandstone, porous"
74.32905484247375,84.13068844807468,"Limestone, porous"
84.13068844807468,101.28354725787631,"Marl/shale, seal"
101.28354725787631,164.17736289381563,"Limestone, porous"
164.17736289381563,189.08984830805136,"Sandstone, porous"
189.08984830805136,201.34189031505252,"Limestone, compact"
201.34189031505252,300.1750291715286,"Limestone, porous"
300.1750291715286,322.6371061843641,"Dolomite, compact"
322.6371061843641,524.7957992998834,"Limestone, compact"
524.7957992998834,551.3418903150525,"Limestone, porous"
551.3418903150525,615.0525087514585,"Limestone, compact"
615.0525087514585,643.6406067677947,"Marl/shale, seal"
643.6406067677947,659.5682613768962,"Sandstone, porous"
659.5682613768962,682.8471411901984,"Sandstone, compact"
682.8471411901984,742.8821470245041,"Sandstone, porous"
742.8821470245041,818.8448074679113,"Marl/shale, seal"
818.8448074679113,852.7421236872813,"Sandstone, porous"
852.7421236872813,883.3722287047842,"Marl/shale, seal"
883.3722287047842,905.8343057176196,"Sandstone, porous"
905.8343057176196,921.3535589264877,"Marl/shale, seal"
921.3535589264877,972.4037339556593,"Sandstone, porous"
972.4037339556593,1056.534422403734,"Limestone, compact"
1056.534422403734,1084.7141190198365,"Marl/shale, seal"
1084.7141190198365,1103.5005834305716,"Limestone, compact"
1103.5005834305716,1118.6114352392067,"Marl/shale, seal"
1118.6114352392067,1124.7374562427071,"Marl/shale, source rock"
1124.7374562427071,1162.7187864644109,"Limestone, compact"
1162.7187864644109,1191.7152858809802,"Marl/shale, seal"
1191.7152858809802,1214.1773628938156,"Limestone, compact"
1214.1773628938156,1253.3838973162194,"Marl/shale, seal"
1253.3838973162194,1378.7631271878647,"Limestone, compact"
1378.7631271878647,1386.931155192532,"Marl/shale, seal"
1386.931155192532,1414.7024504084013,"Limestone, compact"
1420.0116686114352,1602.5670945157526,"Limestone, compact"
1602.5670945157526,1688.3313885647608,"Marl/shale, seal"
1688.3313885647608,1702.6254375729288,"Limestone, compact"
1702.6254375729288,1731.213535589265,"Limestone, porous"
1731.213535589265,1737.7479579929989,"Marl/shale, seal"
1737.7479579929989,1748.7747957993,"Limestone, compact"
1748.7747957993,1774.9124854142358,"Marl/shale, seal"
1774.9124854142358,1793.2905484247374,"Limestone, compact"
1793.2905484247374,1827.5962660443406,"Marl/shale, seal"
1827.5962660443406,1840.2567094515753,"Limestone, compact"
1840.2567094515753,1861.4935822637106,"Marl/shale, seal"
1861.4935822637106,1871.2952158693115,"Limestone, compact"
1871.2952158693115,2000.350058343057,"Marl/shale, seal"
2000.350058343057,2016.6861143523922,"Marl/shale, source rock"
2016.6861143523922,2262.9521586931155,"Marl/shale, seal"
2262.9521586931155,2444.2823803967326,"Sandstone, porous"
2444.2823803967326,2450.0,"Limestone, compact"

In [12]:

print(strip.to_las3())

~Lithology_Parameter
LITH .   Striplog         : Lithology source          {S}
LITHD.   MD               : Lithology depth reference {S}

~Lithology_Definition
LITHT.M                   : Lithology top depth       {F}
LITHB.M                   : Lithology base depth      {F}
LITHD.                    : Lithology description     {S}

~Lithology_Data | Lithology_Definition
7.7596266044340725,32.67211201866978,"Limestone, porous"
32.67211201866978,74.32905484247375,"Sandstone, porous"
74.32905484247375,84.13068844807468,"Limestone, porous"
84.13068844807468,101.28354725787631,"Marl/shale, seal"
101.28354725787631,164.17736289381563,"Limestone, porous"
164.17736289381563,189.08984830805136,"Sandstone, porous"
189.08984830805136,201.34189031505252,"Limestone, compact"
201.34189031505252,300.1750291715286,"Limestone, porous"
300.1750291715286,322.6371061843641,"Dolomite, compact"
322.6371061843641,524.7957992998834,"Limestone, compact"
524.7957992998834,551.3418903150525,"Limestone, porous"
551.3418903150525,615.0525087514585,"Limestone, compact"
615.0525087514585,643.6406067677947,"Marl/shale, seal"
643.6406067677947,659.5682613768962,"Sandstone, porous"
659.5682613768962,682.8471411901984,"Sandstone, compact"
682.8471411901984,742.8821470245041,"Sandstone, porous"
742.8821470245041,818.8448074679113,"Marl/shale, seal"
818.8448074679113,852.7421236872813,"Sandstone, porous"
852.7421236872813,883.3722287047842,"Marl/shale, seal"
883.3722287047842,905.8343057176196,"Sandstone, porous"
905.8343057176196,921.3535589264877,"Marl/shale, seal"
921.3535589264877,972.4037339556593,"Sandstone, porous"
972.4037339556593,1056.534422403734,"Limestone, compact"
1056.534422403734,1084.7141190198365,"Marl/shale, seal"
1084.7141190198365,1103.5005834305716,"Limestone, compact"
1103.5005834305716,1118.6114352392067,"Marl/shale, seal"
1118.6114352392067,1124.7374562427071,"Marl/shale, source rock"
1124.7374562427071,1162.7187864644109,"Limestone, compact"
1162.7187864644109,1191.7152858809802,"Marl/shale, seal"
1191.7152858809802,1214.1773628938156,"Limestone, compact"
1214.1773628938156,1253.3838973162194,"Marl/shale, seal"
1253.3838973162194,1378.7631271878647,"Limestone, compact"
1378.7631271878647,1386.931155192532,"Marl/shale, seal"
1386.931155192532,1414.7024504084013,"Limestone, compact"
1420.0116686114352,1602.5670945157526,"Limestone, compact"
1602.5670945157526,1688.3313885647608,"Marl/shale, seal"
1688.3313885647608,1702.6254375729288,"Limestone, compact"
1702.6254375729288,1731.213535589265,"Limestone, porous"
1731.213535589265,1737.7479579929989,"Marl/shale, seal"
1737.7479579929989,1748.7747957993,"Limestone, compact"
1748.7747957993,1774.9124854142358,"Marl/shale, seal"
1774.9124854142358,1793.2905484247374,"Limestone, compact"
1793.2905484247374,1827.5962660443406,"Marl/shale, seal"
1827.5962660443406,1840.2567094515753,"Limestone, compact"
1840.2567094515753,1861.4935822637106,"Marl/shale, seal"
1861.4935822637106,1871.2952158693115,"Limestone, compact"
1871.2952158693115,2000.350058343057,"Marl/shale, seal"
2000.350058343057,2016.6861143523922,"Marl/shale, source rock"
2016.6861143523922,2262.9521586931155,"Marl/shale, seal"
2262.9521586931155,2444.2823803967326,"Sandstone, porous"
2444.2823803967326,2450.0,"Limestone, compact"

ASIDE¶

$$ x + 2y - 8 = 0 $$

$$ -\frac{1}{2}x + 4 = y $$

Let's look at this through the lens of linear algebra:

$$ Gm = d $$

and machine learning:

$$ Xm = y $$

In fact, re-reading this myself, I think it would have made more sense to talk about a model like:

$$ ax + by = c $$

is equivalent to:

$$ \frac{a}{b}x + \frac{c}{b} = y $$

Note that $y$ is different here, it's the 'target', not another variable.

Then we have some data, $\mathbf{X} = [x_1, x_2, x_3]$ and some coefficients $a, b, c$.

In [6]:

X = [
  [0, 1],
  [8, 1],
]

y = [4, 0]

In [10]:

X = [
  [8, 1],
]

y = [0]

In [12]:

import numpy.linalg as la

la.lstsq(X, y)

<ipython-input-12-0d9c18a3ef88>:3: FutureWarning: `rcond` parameter will change to the default of machine precision times ``max(M, N)`` where M and N are the input matrix dimensions.
To use the future default and silence this warning we advise to pass `rcond=None`, to keep using the old, explicitly pass `rcond=-1`.
  la.lstsq(X, y)

Out[12]:

(array([-0.,  0.]), array([], dtype=float64), 1, array([8.06225775]))

In [ ]:

Striplog from image¶

Make a legend¶

Making a striplog¶

Write CSV or LAS3¶

ASIDE¶

Making a `striplog`¶