A package to compute a total of 261 texture features from voxel lists.
- 11/21/2018. An option to skip tests is implemented. By default, tests are skipped. If you do want to test, set 'do_test' variable to True at the top of the code.
- 9/9/2018. ZLNU is renamed to ZSNU standing for zone-size non-uniformity.
- 9/9/2018. Test codes are added for robustness.
-
If you are not at https://github.com/metavol/ptexture, please go there.
-
Click the green button 'Clone or download' and then click 'Download ZIP'
-
Save the zip on your computer anywhere
-
Extract the zip
-
Create a folder such as 'my_data' on the top level of extracted folder (same level as testdata1, testdata2)
-
Copy your voxel data ('case1.txt' for example) to the above created folder.
-
Run Jupyter Notebook or Jupyter Lab
-
Open 'ptexture_with_text.ipynb'
-
'Edit' menu -> 'Clear All Outputs'
-
Go to the bottom of the source code
-
Find the this line in the bottom box:
df = superBatch(directory = r'testdata1', dim = 3, nbin = 64, lo = 0, hi = 20, connection = 26, ngldm_mode = 1, full = False, report_progress = True, column_renamer = None)
-
Change 'testdata1' to your folder name that was created in 5)
-
'Run' menu -> 'Run All Cells'
-
If successfully done, an excel file is created.
If any errors occure, please contact Kenji Hirata.
Kenji Hirata, MD, PhD
Hokkaido University, Sapporo, Japan