Forecast Future Dataedit

We can use the forecast command to generate future data points. Again, the -s flag saves data to the default bucket, and facilitates data vizualization. Response is written to standard output if the -s flag is omitted.

loudml -e "forecast-model cpu_utilization_asg_misconfiguration -f now-5m -t now+6h"

Output data points contain the following information:

  • @mean_value: the original data point value or null
  • mean_value: the predicted normal value named according to the JSON model definition
  • lower_mean_value: the minimum normal value with 68 percent confidence.
  • upper_mean_value: the maximum normal value with 68 percent confidence

The output should look like this:

timestamp          @mean_value         loudml.mean_value
1569180000.0       31.829             29.897
1569180300.0       33.161             32.23
1569180600.0       31.829             30.063
1569180900.0       33.203             31.157
1569181200.0       34.473             33.131