PythonicDISORT 1.8__tar.gz → 1.9__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: PythonicDISORT
3
- Version: 1.8
3
+ Version: 1.9
4
4
  Summary: Discrete Ordinates Solver for the (1D) Radiative Transfer Equation in a single or multi-layer atmosphere.
5
5
  Author-email: Dion HO Jia Xu <dh3065@columbia.edu>
6
6
  Project-URL: Homepage, https://github.com/LDEO-CREW/Pythonic-DISORT
@@ -74,7 +74,7 @@ These notebooks double up as examples of how to use PythonicDISORT. The tests wh
74
74
  (no corresponding Jupyter Notebook)
75
75
  * Test Problem 11: Single-Layer vs. Multiple Layers
76
76
  * Test Problem I: Antiderivative / integration functionality
77
- * Test Problem Interp: Interpolation + NT Correction functionality
77
+ * Test Problem NT: NT Correction + Interpolation functionality
78
78
 
79
79
  # Installation
80
80
 
@@ -100,7 +100,14 @@ is required to properly run the last section (section 6).
100
100
  ## Compatibility
101
101
 
102
102
  The PythonicDISORT package should be system agnostic given its minimal dependencies and pure Python code.
103
- Everything in the repository was built and tested on Windows 11.
103
+ Everything in the repository was built and tested on Windows 11. It is written and tuned for double-precision.
104
+
105
+ **Known issue:** if SciPy uses Intel MKL 2024.2.x (check with `conda list mkl`), Python may crash
106
+ when `NQuad >= 44` and `NLayers >= use_banded_solver_NLayers` (default 10).
107
+ This is a bug in MKL's banded solver `dgbsv`, which `scipy.linalg.solve_banded` calls, and we observed it on Windows 11 with an AMD CPU.
108
+ To avoid it, either set the environment variable `MKL_CBWR=COMPATIBLE` before starting Python
109
+ (for a conda environment: `conda env config vars set MKL_CBWR=COMPATIBLE`, then reactivate),
110
+ or install an unaffected MKL version (2024.1.0 and 2025.3.0 worked in our tests) or OpenBLAS.
104
111
 
105
112
  # Acknowledgements
106
113
 
@@ -52,7 +52,7 @@ These notebooks double up as examples of how to use PythonicDISORT. The tests wh
52
52
  (no corresponding Jupyter Notebook)
53
53
  * Test Problem 11: Single-Layer vs. Multiple Layers
54
54
  * Test Problem I: Antiderivative / integration functionality
55
- * Test Problem Interp: Interpolation + NT Correction functionality
55
+ * Test Problem NT: NT Correction + Interpolation functionality
56
56
 
57
57
  # Installation
58
58
 
@@ -78,7 +78,14 @@ is required to properly run the last section (section 6).
78
78
  ## Compatibility
79
79
 
80
80
  The PythonicDISORT package should be system agnostic given its minimal dependencies and pure Python code.
81
- Everything in the repository was built and tested on Windows 11.
81
+ Everything in the repository was built and tested on Windows 11. It is written and tuned for double-precision.
82
+
83
+ **Known issue:** if SciPy uses Intel MKL 2024.2.x (check with `conda list mkl`), Python may crash
84
+ when `NQuad >= 44` and `NLayers >= use_banded_solver_NLayers` (default 10).
85
+ This is a bug in MKL's banded solver `dgbsv`, which `scipy.linalg.solve_banded` calls, and we observed it on Windows 11 with an AMD CPU.
86
+ To avoid it, either set the environment variable `MKL_CBWR=COMPATIBLE` before starting Python
87
+ (for a conda environment: `conda env config vars set MKL_CBWR=COMPATIBLE`, then reactivate),
88
+ or install an unaffected MKL version (2024.1.0 and 2025.3.0 worked in our tests) or OpenBLAS.
82
89
 
83
90
  # Acknowledgements
84
91
 
@@ -8,7 +8,7 @@ dependencies = [
8
8
  "scipy>=1.8.0",
9
9
  ]
10
10
  name = "PythonicDISORT"
11
- version = "1.8"
11
+ version = "1.9"
12
12
  authors = [
13
13
  { name="Dion HO Jia Xu", email="dh3065@columbia.edu" },
14
14
  ]