SaS Gurus, I work in a mineral processing facility and I'm experimenting around with SaS Enterprise Miner to see if I can determine which sensor states and sequences lead to certain types of alarms. I'm really after some advice as to the best preparation for the data. As a starting point, can I analyse it in its current form (see below), or do I need to convert it into timeseries data every 100 ms. The data is stored from our sensors only when the value changes. It is stored in the following timestamped format: <Date Time>, <Variable Name>, <Variable Value>. Example Below. TimeStamp Variable Value 3/02/2014 16:02:34.1 Position 22.34 3/02/2014 16:02:34.3 Position 22.37 3/02/2014 16:03:32.7 Speed 83 3/02/2014 16:04:01.6 Position 22.39 3/02/2014 16:04:03.2 Speed 77 3/02/2014 16:07:29.0 Position 22.48 While the alarm instances are just identified by the occurrence timestamp. TimeStamp Alarm 3/02/2014 16:03:23.0 Alarm 1 3/02/2014 18:01:02.0 Alarm 1 Any advice would be greatly appreciated. James Perth, Australia ... View more