You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
which cause csvlook think there are more columns in the second line.:
agate/table/from_csv.py:74: RuntimeWarning: Error sniffing CSV dialect: Could not determine delimiter
ValueError: Row 0 has 17 values, but Table only has 15 columns.
Could this be changed a bit, like replacing x+yi ?
Library context
Software
version
rocblas
ROCm-5.1.3
The text was updated successfully, but these errors were encountered:
Thanks @littlewu2508 for your suggestion, I agree that convention makes parsers more trouble than it is worth. The problem is we may already have this handled so need to update our scripts as well, will review and get back to you.
The complex output for rocblas-bench will be "(real: maginary)" after develop branch commit 79fcb2b
Output elsewhere has not change to minimize any effects on other parsers. If this causes no trouble we can consider changing to this format for all complex number output sometime in the future.
Is your feature request related to a problem? Please describe.
I use rocblas-bench combined with csvlook to get a visualized result, for example:
However when benchmarking complex floating point operations, the results use (x,y) to express complex number:
which cause csvlook think there are more columns in the second line.:
Could this be changed a bit, like replacing x+yi ?
Library context
The text was updated successfully, but these errors were encountered: