the data goes on for 24 hours... is there a simpler way to resample minutely data into hourly data?... the other coulm being averaged or interpolated... similar to pandas python resample ?
you would need a string to date node to parse the format into the knime date format.
Than you use the Date to String node to extract an hourly format only (e.g. yyyy-MM-dd'T'HH)
Finally you can use a groupby node to get your sample. Groupby the last column generated by the time to string node and select an average in your value column.
Given that i have an hourly temperature data of a city, how would i convert into 15minute (quarterly) frequency data... using linear or cubic interpolation or any other method ?
You could use Constant Value Column to add a column containing the integer value 4, then One Row to Many to make four copies of every row, then a Moving Average node to do the interpolation of the temperature values - various kinds of interpolation are available.
If you need the interpolated 15-minute time points as well, use Time Series Generator set to Saw Tooth Wave, height 45, length 4 and mean 22.5; Double to Int to convert the generated time series to integer; then Date Time Shift using the integer time series column as the shift value in minutes and the original time column as the date reference.
Tom, I had a similar problem. I had minute time series measurements for a device which I needed to downsample to ten minute resolution. To do this I did the following -
Sorted by device and time
Extracted the minute from the date and time
Started a group loop based on the device
Lagged the measurement column by ten, giving me ten columns, going back ten minutes
Calculate the average for those ten columns (i.e.: a ten minute average)
Calculate the modulus of the minute column (i.e.: 30 mod 10)
Used a rule-based row filter to select only those rows where the modulus was zero, thus being the ten minute multiple