|
| 1 | +### June 13, 2023 |
| 2 | ++ Python podcasts suggestion from Tomasz: [PythonBytes](https://pythonbytes.fm/) |
| 3 | ++ David talked about an new Python module called “Pandas AI” that did find useful if you have a paid chatGPT account |
| 4 | + + https://github.com/gventuri/pandas-ai |
| 5 | + + “Pandas AI is a Python library that integrates generative artificial intelligence capabilities into Pandas, making dataframes conversational” |
| 6 | + + David id also find a poorly written blog post that was claiming featured that Pandas AI does NOT have, so stay away from this article… |
| 7 | + + https://levelup.gitconnected.com/introducing-pandasai-the-generative-ai-python-library-568a971af014 |
| 8 | ++ We talked about when we have used chatGPT to write some Python code snippets, and what were our results. |
| 9 | + + The results were mostly positive, but we talked about the benefits of already knowing Python well enough to formulate request more precisely and evaluate how well the chatGPT responses were |
| 10 | + + Someone mentioned that chatGPT has become as an alternative to StackOverflow, specially if you are in a hurry |
| 11 | + + Someone mentioned Github Copilot: “Those of us who have GitHub educator accounts have free access to Copilot. Have not tried it. Very reluctant, personally.” Which uses AI to write code for you. |
| 12 | + + https://en.wikipedia.org/wiki/GitHub_Copilot |
| 13 | + + As a counter argument there is this article [“Why I don’t use Copilot”](https://inkdroid.org/2023/06/04/copilot/) |
| 14 | + + Will StackOverflow become obsolete with the revolution in AI? Yamil thinks that it is a good inspiration for prompts, and still has great information |
| 15 | + + We saw an example of sharing a snippet of object oriented Python code to ask chatGPT to explain what is missing |
| 16 | + + One of the participants was glad to get the explanations from chatGPT of what was missing in their object oriented code |
| 17 | + + Here is the link to the chat https://chat.openai.com/share/fea426fb-cb02-4b38-9f42-128f59115fc4 |
| 18 | + + A recent Code4Lib article that talked about using AI generated code was shared [“Utilizing R and Python for Institutional Repository Daily Jobs”](https://journal.code4lib.org/articles/17134) |
| 19 | + + We briefly talked about the ethics of using AI written code that was trained on code that other published publicly on Github, but without their explicit consent |
| 20 | + + Podcast example crated by AI: |
| 21 | + “I’ve been listening to this series in the Planet Money podcast where they try to make an entire podcast episode made by AI:” https://www.npr.org/series/1178395718/planet-money-makes-an-episode-using-ai |
| 22 | + + Charles asked if anyone was using Python to automate work with the Azure cloud computing platform |
| 23 | + + https://en.wikipedia.org/wiki/Microsoft_Azure |
| 24 | + + We briefly talked about [“Azure Functions”](https://learn.microsoft.com/en-us/azure/azure-functions/functions-overview?pivots=programming-language-csharp) which seem similar to [AWS Lambda](https://en.wikipedia.org/wiki/AWS_Lambda) |
| 25 | ++ We talked about a great site and free book that many people use to get started with Python [“Python for Everyone”](https://www.py4e.com/) |
| 26 | ++ We also talked about the well known and still very popular Python [Requests](https://requests.readthedocs.io/en/latest/) module, and but also the newer and “async compatible” [HTTPX](https://www.python-httpx.org/) module, which was also mentioned on the Python Slack channel. |
| 27 | + |
1 | 28 | ### May 30, 2023 |
2 | 29 | + David shared his code utlizing `pymarc` to harvest and clean OCLC records. An older example of code: https://github.com/derlandson/PyCat |
3 | 30 | + Demo of Match MARC toolset as well. |
|
0 commit comments