sensortoolkit.calculate._cv

This module computes the coefficient of variation (CV), a measure of precision, for collocated sensors recording data concurrently.

Note

CV as defined by U.S. EPA’s Performance Targets Reports is measured for periods where all sensors in the evaluation group are measuring concurrently.

For example, if three sensors are included in a deployment group, the CV will only be calculated for periods where all three sensors are operating normally and recording values simultaneously.

U.S. EPA’s Performance Targets Reports calculate CV as

\[CV = \frac{SD}{\bar{x}}\times 100\]

where \(\bar{x}\) is the deployment averaged sensor concentration for a field test, and \(SD\), the standard deviation, is defined as

\[SD = \sqrt{\frac{1}{(N\times M)-1}\sum_{j=1}^{M}\left[ \sum_{i=1}^{N}(x_{ij} - \bar{x_i})^2\right]}\]

and where:

\(M\) = number of identical sensors operated simultaneously during a field test

\(N\) = number of measurement intervals during which all identical instruments are operating and returning valid averages over the duration of the field test

\(x_{ij}\) = Sensor concentration for measurement interval \(i\) and sensor unit \(j\).

\(\bar{x_i}\) = Intersensor average sensor concentration for measurement interval \(i\). All sensor units deployed for testing must have recorded non-null values for measurement interval \(i\) to compute \(\bar{x_i}\).


@Author:
Samuel Frederick, NSSC Contractor (ORAU)
U.S. EPA / ORD / CEMM / AMCD / SFSB
Created:

Wed Jan 29 10:03:27 2020

Last Updated:

Tue Jul 13 09:21:40 2021

Functions

cv

Compute CV for set of sensor dataframes and indicated parameter.