notebook.community

Edit and run



In [130]:

    
import pandas as pd
import zipfile; import os; import urllib
import lasio; import matplotlib.pyplot as plt; import welly; from welly import Well
% matplotlib inline



In [18]:

    
## Download and unzip databases from KGS to a temp folder
os.makedirs('./temp', exist_ok=True)
raspath = 'http://www.kgs.ku.edu/PRS/Ora_Archive/ks_elog_scans.zip'
laspath = 'http://www.kgs.ku.edu/PRS/Ora_Archive/ks_las_files.zip'

ras_handle, _ = urllib.request.urlretrieve(raspath)
las_handle, _ = urllib.request.urlretrieve(laspath)

with zipfile.ZipFile(ras_handle,'r') as zip_ras:
    zip_ras.extractall('temp')
with zipfile.ZipFile(las_handle,'r') as zip_las:
    zip_las.extractall('temp')

Elogs is collection of KGS TIFF files; lases is KGS .las files



In [57]:

    
elogs = pd.read_csv('temp/ks_elog_scans.txt', parse_dates=True)
lases = pd.read_csv('temp/ks_las_files.txt', parse_dates=True)



In [83]:

    
elogs_mask = elogs['KID'].isin(lases['KGS_ID'])  # Create mask for elogs
both_elog = elogs[elogs_mask] # select items elog that fall in both
both_elog.drop_duplicates('KID') # remove duplicates
print('How many logs fall in both and have unique KGS_ID? '+str(both_elog_unique.shape[0]))
both_elog_unique_new = both_elog_unique.loc['2000-1-1' : '2017-1-1']
both_elog_unique_new;









    



How many logs fall in both and have unique KGS_ID? 14289



In [84]:

    
lases_mask = lases['KGS_ID'].isin(elogs['KID'])  # Create mask for elogs
both_lases = lases[las_mask] # select items elog that fall in both
both_lases_unique = both_lases.drop_duplicates('KGS_ID') # remove duplicates
print('Other direction -- how many logs fall in both and have unique KGS_ID? '+str(both_lases_unique.shape[0]))
if both_elog_unique.shape[0] == both_lases_unique.shape[0]:
    print('Same in both directions.')
both_lases_unique;









    



Other direction -- how many logs fall in both and have unique KGS_ID? 14289
Same in both directions.

Download a random las file, unzip, and render with lasio



In [108]:

    
# Each time you run this you'll get a new las file, so...
randomlas_line = both_lases_unique.loc['2000-1-1' : '2017-1-1'].sample(n=1)



In [109]:

    
str(randomlas_line['URL'])
randomlas_line









    Out[109]:







  
    
      
      KGS_ID
      Latitude
      Longitude
      Location
      Operator
      Lease
      API
      Elevation
      Elev_Ref
      Depth_start
      Depth_stop
      URL
    
  
  
    
      2013
      1044186387
      37.405423
      -95.67953
      T30S R16E, Sec. 29,  SE NE NE SW
      A X & P, Inc.
      Unit 1 - Wolfe West WW 29-H
      15-205-28064
      803.0
      KB
      -5.25
      842.0
      http://www.kgs.ku.edu/WellLogs/kcc_logs_2014/1...



In [110]:

    
randomlas_url = randomlas_line['URL'].tolist()[0]
randomlas_dl, _ = urllib.request.urlretrieve(randomlas_url)

with zipfile.ZipFile(randomlas_dl,'r') as zip_randomlas:
    zip_randomlas.extractall('temp')



In [111]:

    
randomlas_line['KGS_ID'].tolist()[0]









    Out[111]:





1044186387

Read in well log and display using welly



In [144]:

    
l = Well.from_las('temp/1044249438.las') # hardcoded until I figure out how to get filename
l.data['GR'];
# l.plot(tracks = [['GR','NEU']]);
l.data['GR'].plot(lw=0.5)

Next steps:

Connect tiff files (format) with log curves from lases -- will need to use welly dictionary (?)
Create function to download a random subset of paired las / tiff files
Parameters: Number of files to download; date range
Returns: consistently named, paired training data