Indicator Krigging

In [1]:

#points

#points

In [2]:

from volumetricspy.stats import Dot,CloudPoints, Spherical, IndicatorOridinaryKrigging
import pandas as pd
import numpy as np
import geopandas as gpd
import matplotlib.pyplot as plt
import seaborn as sns
from scipy.spatial import Voronoi, voronoi_plot_2d

from volumetricspy.stats import Dot,CloudPoints, Spherical, IndicatorOridinaryKrigging import pandas as pd import numpy as np import geopandas as gpd import matplotlib.pyplot as plt import seaborn as sns from scipy.spatial import Voronoi, voronoi_plot_2d

In [3]:

p1 = Dot(x=50,y=40, fields={'facies':'sand'})
p2 = Dot(x=20,y=90, fields={'facies':'coal'})
p3 = Dot(x=60,y=70, fields={'facies':'shale'})
p = CloudPoints(points = [p1,p2,p3])
#p.one_hot_encode('facies')
print(p.df())

p1 = Dot(x=50,y=40, fields={'facies':'sand'}) p2 = Dot(x=20,y=90, fields={'facies':'coal'}) p3 = Dot(x=60,y=70, fields={'facies':'shale'}) p = CloudPoints(points = [p1,p2,p3]) #p.one_hot_encode('facies') print(p.df())

      x     y     z   crs facies
0  50.0  40.0  None  None   sand
1  20.0  90.0  None  None   coal
2  60.0  70.0  None  None  shale

In [4]:

up1 = Dot(x=70,y=27)
up = CloudPoints(points = [up1])
up.df()

up1 = Dot(x=70,y=27) up = CloudPoints(points = [up1]) up.df()

Out[4]:

	x	y	z	crs
0	70.0	27.0	None	None

In [5]:

s1 = Spherical(
    sill = 1.,
    range = 330,
    nuggets = 0.
)

s1.plot(np.linspace(0,500,50))

s1 = Spherical( sill = 1., range = 330, nuggets = 0. ) s1.plot(np.linspace(0,500,50))

Out[5]:

<AxesSubplot:>

In [6]:

p.df()

p.df()

Out[6]:

	x	y	z	crs	facies
0	50.0	40.0	None	None	sand
1	20.0	90.0	None	None	coal
2	60.0	70.0	None	None	shale

In [7]:

ik = IndicatorOridinaryKrigging(
    known_cp = p,
    unknown_cp = up,
    variogram_model = s1
)
ik

ik = IndicatorOridinaryKrigging( known_cp = p, unknown_cp = up, variogram_model = s1 ) ik

Out[7]:

IndicatorOridinaryKrigging(known_cp=CloudPoints(points=[Dot(x=50.0, y=40.0, z=None, crs=None, fields={'facies': 'sand'}), Dot(x=20.0, y=90.0, z=None, crs=None, fields={'facies': 'coal'}), Dot(x=60.0, y=70.0, z=None, crs=None, fields={'facies': 'shale'})]), unknown_cp=CloudPoints(points=[Dot(x=70.0, y=27.0, z=None, crs=None, fields=None)]), variogram_model=Spherical(sill=1.0, nugget=0.0, range=330.0))

In [8]:

ik.variogram_model.plot(np.linspace(0,500,50))

ik.variogram_model.plot(np.linspace(0,500,50))

Out[8]:

<AxesSubplot:>

In [9]:

rcp = ik.forward('facies')
rcp.df().info()

rcp = ik.forward('facies') rcp.df().info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1 entries, 0 to 0
Data columns (total 11 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   x               1 non-null      float64
 1   y               1 non-null      float64
 2   z               0 non-null      object 
 3   crs             0 non-null      object 
 4   sand            1 non-null      float64
 5   sand_variance   1 non-null      float64
 6   coal            1 non-null      float64
 7   coal_variance   1 non-null      float64
 8   shale           1 non-null      float64
 9   shale_variance  1 non-null      float64
 10  facies          1 non-null      object 
dtypes: float64(8), object(3)
memory usage: 216.0+ bytes

In [10]:

#df = pd.read_csv('sample_data_biased.csv')     # load our data table
df = pd.read_csv('https://raw.githubusercontent.com/GeostatsGuy/GeoDataSets/master/sample_data_biased.csv')
df['Facies'] = df['Facies'].map({0:'sand',1:'shale'})
df

#df = pd.read_csv('sample_data_biased.csv') # load our data table df = pd.read_csv('https://raw.githubusercontent.com/GeostatsGuy/GeoDataSets/master/sample_data_biased.csv') df['Facies'] = df['Facies'].map({0:'sand',1:'shale'}) df

Out[10]:

	X	Y	Facies	Porosity	Perm
0	100	900	shale	0.115359	5.736104
1	100	800	shale	0.136425	17.211462
2	100	600	shale	0.135810	43.724752
3	100	500	sand	0.094414	1.609942
4	100	100	sand	0.113049	10.886001
...	...	...	...	...	...
284	190	199	shale	0.109411	3.247265
285	190	419	sand	0.099541	1.528003
286	920	379	sand	0.094099	2.514597
287	60	149	sand	0.100167	3.636854
288	390	549	shale	0.153335	165.699353

289 rows × 5 columns

In [11]:

cp = CloudPoints().from_df(df, x='X',y='Y', fields=['Facies','Porosity','Perm'])
cp.df()

cp = CloudPoints().from_df(df, x='X',y='Y', fields=['Facies','Porosity','Perm']) cp.df()

Out[11]:

	x	y	z	crs	Facies	Porosity	Perm
0	100.0	900.0	None	None	shale	0.115359	5.736104
1	100.0	800.0	None	None	shale	0.136425	17.211462
2	100.0	600.0	None	None	shale	0.135810	43.724752
3	100.0	500.0	None	None	sand	0.094414	1.609942
4	100.0	100.0	None	None	sand	0.113049	10.886001
...	...	...	...	...	...	...	...
284	190.0	199.0	None	None	shale	0.109411	3.247265
285	190.0	419.0	None	None	sand	0.099541	1.528003
286	920.0	379.0	None	None	sand	0.094099	2.514597
287	60.0	149.0	None	None	sand	0.100167	3.636854
288	390.0	549.0	None	None	shale	0.153335	165.699353

289 rows × 7 columns

In [12]:

cp.plot(hue='Facies')

cp.plot(hue='Facies')

Out[12]:

<AxesSubplot:xlabel='x', ylabel='y'>

In [13]:

s2 = Spherical(
    sill = 1.,
    range = 150,
    nuggets = 0.
)

s2.plot(np.linspace(0,500,50))

s2 = Spherical( sill = 1., range = 150, nuggets = 0. ) s2.plot(np.linspace(0,500,50))

Out[13]:

<AxesSubplot:>

In [14]:

xn = np.linspace(0,1000,50)
yn = np.linspace(0,1000,50)

xx, yy = np.meshgrid(xn, yn)

dfun = pd.DataFrame({"x":xx.flatten(order='F'),"y":yy.flatten(order='F')})
grid = CloudPoints().from_df(dfun, x='x',y='y')

xn = np.linspace(0,1000,50) yn = np.linspace(0,1000,50) xx, yy = np.meshgrid(xn, yn) dfun = pd.DataFrame({"x":xx.flatten(order='F'),"y":yy.flatten(order='F')}) grid = CloudPoints().from_df(dfun, x='x',y='y')

In [15]:

ik2 = IndicatorOridinaryKrigging(
    known_cp = cp,
    unknown_cp = grid,
    variogram_model = s2
)

ik2 = IndicatorOridinaryKrigging( known_cp = cp, unknown_cp = grid, variogram_model = s2 )

In [20]:

rcp2 = ik2.forward('Facies', argmax=True)
rcp2.df()

rcp2 = ik2.forward('Facies', argmax=True) rcp2.df()

Out[20]:

	x	y	z	crs	shale	shale_variance	sand	sand_variance	Facies
0	0.0	0.000000	None	None	0.428006	0.706652	0.571994	0.706652	sand
1	0.0	20.408163	None	None	0.309545	0.550432	0.690455	0.550432	sand
2	0.0	40.816327	None	None	0.197713	0.422455	0.802287	0.422455	sand
3	0.0	61.224490	None	None	0.120137	0.330105	0.879863	0.330105	sand
4	0.0	81.632653	None	None	0.070710	0.279478	0.929290	0.279478	sand
...	...	...	...	...	...	...	...	...	...
2495	1000.0	918.367347	None	None	0.375504	0.677546	0.624496	0.677546	sand
2496	1000.0	938.775510	None	None	0.229328	0.473992	0.770672	0.473992	sand
2497	1000.0	959.183673	None	None	0.114166	0.288207	0.885834	0.288207	sand
2498	1000.0	979.591837	None	None	0.083206	0.186104	0.916794	0.186104	sand
2499	1000.0	1000.000000	None	None	0.173995	0.403432	0.826005	0.403432	sand

2500 rows × 9 columns

In [21]:

rcp2.plot(hue='Facies')

rcp2.plot(hue='Facies')

Out[21]:

<AxesSubplot:xlabel='x', ylabel='y'>

In [19]:

cp.plot(hue='Facies', hue_order=['sand','shale'])

cp.plot(hue='Facies', hue_order=['sand','shale'])

Out[19]:

<AxesSubplot:xlabel='x', ylabel='y'>

In [ ]:

dft = rcp2.df()
dft

dft = rcp2.df() dft

In [ ]:

from numpy.random import default_rng

rng = default_rng(seed=None)

rng.choice(dft['Facies'].unique(), size=10)

from numpy.random import default_rng rng = default_rng(seed=None) rng.choice(dft['Facies'].unique(), size=10)

In [ ]:

for i, r in dft.head().iterrows():
    print(r[['sand','shale']])

for i, r in dft.head().iterrows(): print(r[['sand','shale']])

In [ ]:

def softmax(x):
    e = np.exp(x)
    return e / e.sum()

def softmax(x): e = np.exp(x) return e / e.sum()

In [ ]:

softmax([-0.2,0.2,0.6,0.4])

softmax([-0.2,0.2,0.6,0.4])

In [ ]: