ASCEE/pyqtgraph
5
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

avoid numpy scalar overflow

Browse Source 
the "offset" argument passed into rescaleData() is typically an element
of an ndarray, i.e. it comes from the lower bound value of the "levels"
ndarray. as such, arithmetic operations on it can overflow.

this triggers a runtime warning in the test suite.

the workaround is to convert it to a Python (integer) scalar.
This commit is contained in:
KIU Shueng Chuan 2021-04-09 10:08:03 +08:00
parent 0f94c17d86
commit 3be8cafff4
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
pyqtgraph/functions.py
View File

@ -1046,7 +1046,9 @@ def _rescaleData_nditer(data_in, scale, offset, work_dtype, out_dtype, clip):
# casting to an int32 will lose the fractional part, therefore the
# output dtype must be an integer kind.
lim_in = np.iinfo(data_in.dtype)
dst_bounds = scale * (lim_in.min - offset), scale * (lim_in.max - offset)
# convert numpy scalar to python scalar to avoid overflow warnings
lo = offset.item(0) if isinstance(offset, np.number) else offset
dst_bounds = scale * (lim_in.min - lo), scale * (lim_in.max - lo)
if dst_bounds[1] < dst_bounds[0]:
dst_bounds = dst_bounds[1], dst_bounds[0]
lim32 = np.iinfo(np.int32)