data generator

[till-teb]

For data prediction we will need a lot test data, that's why we need the data generator.
The following settings should be possible for the data generator function before he produce the dataset:

• How many columns?
• Min value
• Max value
• distribution of the data
  These settings should be set before then the dataset will generated itself.

---

The DataFrame contains the following columns:

• expense
• amount
• importance
• category
• date
• label for good day or bad day

[l4nz8]

This is a big problem we have, but I also think it would be too much work to build one from scratch. We would first need real data on which we can base artificial data. If we were to work with random data, it would be unrealistic. I'm not sure if we need generated data at all, because I don't see any point in training any model. It gives us no added value in our analysis.

The only thing we need is a data set to see if our program is working properly.