The first step for external processing is to make the the data public. The window will then move to include raw values between 00:01 (inclusive) and 00:31 (exclusive), with the logarithmic average having a timestamp of 00:01, and so on. This means that the raw values between 00:00 (inclusive) and 00:30 (exclusive) will be aggregated and the resulting logarithmic average will have a timestamp of 00:00. The timestamp of each average value will be at the start of the 30-minute window, just like the built-in eagle.io aggregates (which in turn are derived from OPC Unified Architecture specifications). The average calculation will include 30 minutes of raw data, and produce one output every minute. The sample noise data used in the following example will comprise 10 hours of 1-minute values. And by controlling how many anti-log values are included in the average calculation, this can be achieved in a "rolling" way by having a moving window of raw values included in each iteration. In basic terms, an appropriate logarithmic average can be achieved by calculating the anti-log of each decibel value, finding the average of all the anti-log values, and then finally calculating the log of that average. This means that a simple arithmetic average (such as the AVERAGE aggregate which is available via Processing & Logic) is not a suitable way to find the mean value of a set of noise data. Noise data is measured in decibels, which are expressed on a logarithmic scale. The method used comes from this blog post which uses Excel to perform the calculations. This article will describe how to perform this aggregation using a Google sheet. One specific type of aggregation is used to calculate an average for logarithmic values, such as noise data. In these cases, processing data externally in a Google sheet is an option, and the general steps for doing so are covered in this article. Some specialized types of aggregation are not available via Processing & Logic.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |