DEV Community

Arman Tarkhanian
Arman Tarkhanian

Posted on

#4: Lots of Pandas

I’ll start off this post by saying that it’s been hard to work extensively on the game project due to full time work, but I expected that. My own personal work this week has been trying to find some suitable assets for our game considering we have no designers or animators at the moment. My uncle is hiring a project manager so hopefully she will help with the direction going forward. We’ve been doing a lot of research with AI to use for our game, since we want some autonomous learning for each agent for each game. It’s meant to be very realistic.

As for work, a lot of it has been combing through our Salesforce data and making some exploratory correlations and just seeing what kind of data is there. I found an interesting little correlation between day of the week and motor vehicle accidents as far as our leads go in which they’re far less common on Sundays. This may be due to a lower general volume of calls in the first place but I’ll have to look deeper. I used Simple Salesforce to log in with Python code and spun up a Jupyter notebook to play around with all this stuff, converting the SOQL query results to a DataFrame. I might switch over to some other APIs to see what works, such as SalesforcePy and/or Polars, just to get my feet wet and see what else is out there. I’m also looking to draw even more correlations, especially longitudinally. I also want to determine how much of our data is “unclean,” since it’s all human input and inconsistent.

Overall, as I said last week, fairly dry stuff at the moment but it’s keeping me engaged at least. We got two new hires on Tuesday but we haven’t started collaborating yet. I’m told we should be meeting each other and working together starting next week which may be a nice change of pace. Cheers till then!

Top comments (0)