dotnet / dotnet/machinelearning
DataFrame LoadCsv improvements
- Dominant language
- C#
- Stars
- 9.4k
- Forks
- 2k
- Avg merge
- 2d 20h
- Merged PRs (30d)
- 11
Description
_This issue has been moved from [a ticket on Developer Community](https://developercommunity.visualstudio.com/t/DataFrame-LoadCsv-improvements/10714999)._
---
The DataFrame.LoadCsv method could be improved in a number of ways:
- Use double when the precision in the data allows it. Now float is always used when the data is floating point.
- Allow the user to define NaN in floating point columns data. In R code that we are also using, NA is typically used.
- Speed improvements if possible.
- I think there is a problem with defining culture info in the parameter. I have to do:
Thread.CurrentThread.CurrentCulture = new CultureInfo("en-US"); before the call to LoadCsv to make floating point data with decimal points load correctly (in Sweden).
---
### Original Comments
#### Feedback Bot on 8/1/2024, 10:56 PM:
(private comment, text removed)
Contributor guide
Assessment
This issue has not been assessed yet.