Description of problem: Numeric.reshape raises ValueError exception when newshape has more than 1 dimension. Version-Release number of selected component (if applicable): python-numeric-23.7-2.2.1 How reproducible: from Numeric import * x = array([1,2,3,4], Int) y = reshape(x, (2,2)) Actual results: Traceback (most recent call last): File "aa.py", line 3, in ? y = reshape(x, (2,2)) ValueError: total size of new array must be unchanged Expected results: y = [[1 2] [3 4]] Additional info: Same code works fine with numpy-0.9.6-1.fc5. Issue may have been resolved in later release of Numeric (release notes of later versions mention fixed to 64 bit issues). Many fc5 packages depend on python-numeric, so it is not possible to simply remove python-numeric and use numpy.
Fixed in python-numeric-24.2-1.
Bug is still present in RHEL 5. Can we get the patch applied there too?
Please open a separate bug for RHEL issues.