I am now happily ensconced at PyOhio and have spare cycles to check out your issue! I have not yet been able to reproduce your problem after a few tries, however. First, your code does not run when pasted into a Python file — for one thing, the statement "EpDate = float(EpDate)" is trying to access the "EpDate" before assigning it a value for the first time. Another problem is that your code has many symbols that are never defined — but two friendly fluid dynamics engineers at the table suggested that they looked like NumPy names ("double", "arange"), so maybe I am missing some import statements from the top of your program? Finally, the numbers I got out looked nothing like yours; could it be that the transitions you are making to and from NumPy floats are losing some precision somewhere?
Anyway, to move forward I think I'll need your full .py file; making random changes to what you pasted above in an effort to produce working code is not producing anything like your graph. I'll keep watching this issue this evening and tomorrow, and start up the investigation again if I see you attach a file.
Lance —
I am now happily ensconced at PyOhio and have spare cycles to check out your issue! I have not yet been able to reproduce your problem after a few tries, however. First, your code does not run when pasted into a Python file — for one thing, the statement "EpDate = float(EpDate)" is trying to access the "EpDate" before assigning it a value for the first time. Another problem is that your code has many symbols that are never defined — but two friendly fluid dynamics engineers at the table suggested that they looked like NumPy names ("double", "arange"), so maybe I am missing some import statements from the top of your program? Finally, the numbers I got out looked nothing like yours; could it be that the transitions you are making to and from NumPy floats are losing some precision somewhere?
Anyway, to move forward I think I'll need your full .py file; making random changes to what you pasted above in an effort to produce working code is not producing anything like your graph. I'll keep watching this issue this evening and tomorrow, and start up the investigation again if I see you attach a file.