Rewording of NEWS and CRAN comments

Author: mayer79

NEWS.md

@@ -1,16 +1,16 @@
 # kernelshap 0.9.0
 
-### Major bug fix
+### Bug fix
 
-`kernelshap()` used a wrong weighting logic, leading to values slightly off. This has
-been fixed with the help of Prof Mario Wuethrich of ETHZ and Ian Covert and his
-[wonderful Github repo](https://github.com/iancovert/shapley-regression).
-Now, exact Kernel SHAP returns identical values as exact permutation SHAP.
-All variants of `kernelshap()` had been affected by this (exact, sampling, hybrid).
-For models with interactions up to two, the bug had no consequences - which is why
-it went unnoticed.
+With input from Mario Wuethrich and Ian Covert and his [repo](https://github.com/iancovert/shapley-regression), 
+we have fixed a bug in how `kernelshap()` calculates Kernel weights. 
 
-Fixed in [#168](https://github.com/ModelOriented/kernelshap/pull/168).
+- The differences caused by this are typically very small.
+- Models with interactions of order up to two have been unaffected.
+- Exact Kernel SHAP now provides identical results to exact permutation SHAP.
+
+Fixed in [#168](https://github.com/ModelOriented/kernelshap/pull/168), which also has received
+unit tests against Python's "shap".
 
 ### API
 

cran-comments.md

@@ -1,14 +1,13 @@
-# kernelshap 0.9.0 (MAJOR BUG FIX)
+# kernelshap 0.9.0
 
-We have figured out a major bug in the weighting logic of Kernel SHAP.
+We have figured out a bug in the weighting logic of Kernel SHAP.
 
-This update comes with a fix which has been tested also against different
-Python algorithms.
+This update comes with a fix which has been tested against two other implementations.
 
-I know that summer holidays are ahead, and the last release is not long ago, 
-but I would love to see this change on CRAN very soon.
+I am aware that the last release of {kernelshap} is not too long ago, but I still would love to see
+this fixed before the (well-deserved) summer break.
 
-Michael
+Thanks a lot!
 
 ## Checks