Different machines outputting different results

Hi there,

This is a very very very odd situation. Due to my work conditions, I usually work in different machines (desktops, laptops) since all my data is stored in the cloud, and I can easily access them.

Now comes the odd part: I was having some trouble getting outputted the same results from Ansys and pyansys, and I was sure that the problem was my code so I decided to call it a day and work on it after, with a fresh mind. That was when I changed from my working station, a desktop, and started working on the code home on my laptop. Before changing anything in the code, I decided running it one more time before going into the debugging phase, and that’s where I lost my mind: the results were now exactly the same as the ones outputted by Ansys.

Before jumping into conclusions, I checked if there was any difference between the two machines, like Python version, conda version, IPython version, Spyder version and most importantly pyansys version, among others. All the versions were the same. Not satisfied, I asked some colleagues (some with the same versions as mine, others with different) to run the same files to check the outputs, and not surprisingly, all the results were consistent between themselves, but the wrong one (the same as my desktop). The only machine that outputs the same results as the ones from Ansys is my laptop. How can that be?!

I hope you can help me understanding and solving this odd situation.

Here are some additional information regarding the versions:

1. My laptop (same results as Ansys):

• Architecture 64bit
• python 3.7.4
• pyansys 0.40.0
• numpy 1.18.4
• scipy 1.4.1
• IPython 7.6.1
• matplotlib 3.2.1
• scooby 0.5.4
• conda 4.7.12

1. My desktop (consistent with others but different from Ansys):

• Architecture 64bit
• python 3.7.4
• pyansys 0.40.0
• numpy 1.18.4
• scipy 1.4.1
• IPython 7.6.1
• matplotlib 3.2.1
• scooby 0.5.4
• conda 4.7.12

1. Colleagues configs (consistent with others but different from Ansys):

• Architecture 64bit
• python 3.7.4, python 3.7.6
• pyansys 0.40.0
• numpy 1.18.4
• scipy 1.4.1
• IPython 7.6.1, IPython 7.13.0
• matplotlib 3.2.1
• scooby 0.5.4
• conda 4.7.12, conda 4.8.3

PS: I can share all the simulation files by request. For that please provide me your email address.

Many thanks in advance.

[akaszynski]

That’s an interesting bug! Here’s my work address: [e-mail]

[akaszynski]

Hello,

I’m pinging you on this closed issue because GitHub doesn’t have an email address for you.

ANSYS is busy improving and extending pyansys. As an existing user, we are interested in learning about your experience and opinions using this package. Most notably:

• What is your application?
• How does pyansys facilitate your work?
• What new pyansys features or improvements would you find valuable?
• Your feedback will help us prioritize our investments and make pyansys even better.

On behalf of the pyansys development team, I look forward to hearing any feedback from you as it will be used to improve and advance pyansys.

Feel free to respond here to this issue, or email me at [e-mail] Alex Kaszynski