Google Correlation like workflow with a given time series

Hi, I have a log file (time series) with firewall logs, authentication logs (logins) , and antivirus logs

I would like to extract the frequencies of some event types in 1 hour bins and have a correlation matrix to see which kind of event is more correlated with the rest.

How would I achieve that in Knime ? is that possible ?

I am very new to Knime

thanks in advance

Hi @hakandurgut,

You can use the String to Date&Time node to convert your timestamp column to KNIME Date&Time cells, then the Extract Date&Time Fields node to extract hour (and day of month if needed), then group by hour (and day?) and the event type using count as the aggregation method (the GrouBy node). Afterwards you can use either the JavaScript Bar Chart to compare events distribution by hours or use the Linear Correlation node to get the correlation matrix.

Best,
Anna