Hi, thanks for the excellent tutorial! I encountered an issue where my correlation matrix was filled with NaN values. If anyone else is facing a similar problem, here’s how I resolved it: I used the Hepatitis C Prediction Dataset from 2020 on Kaggle. In this dataset, the category for Hepatitis was labeled as "1=Hepatitis." I filtered the data using: hep = data.loc[data.Category == "1=Hepatitis"] Then, I manually dropped the columns with non-numeric values: hep = hep.drop(columns=['Unnamed: 0', 'Category', 'Sex']) Finally, I removed the rows with NaN values with: hep = hep.dropna() Hope this helps anyone who runs into the same issue!
In the dataset I used, the category for Hepatitis was labeled as "1=Hepatitis". So you can filter the data like this: hep = data.loc[data.Category == "1=Hepatitis"]. Then you need to manually drop the columns with non-numeric values: hep = hep.drop(columns=['Unnamed: 0', 'Category', 'Sex']). Also I had to remove the rows with NaN values: hep = hep.dropna(). Hope this helps you!
