Evaluation Objects

sensortoolkit contains several class objects to aid users in various data analysis tasks. These objects can be roughly divided into two categories: testing attribute objects and evaluation objects.

Since users may wish to evaluate multiple sensor types and numerous pollutants, the intent in organizing the evaluation workflow into a set of recognizable and consistent class objects is to provide an easy-to-navigate platform for conducting evaluations and data analysis, while also allowing a high degree of customization.

---

Conducting performance evaluations of air sensors with sensortoolkit is accomplished via the library’s evaluation objects, sensortoolkit.SensorEvaluation and sensortoolkit.PerformanceReport. sensortoolkit.SensorEvaluation can be used to conduct performance evaluations for air sensors with EPA’s recommended performance metrics and targets for either fine particulate matter or ozone sensors, and figures and summary statistics can be generating for reviewing sensor performance. Testing reports utilizing EPA’s reporting templates for base testing of air sensors under NSIM applications are created via the sensortoolkit.PerformanceReport class.

Overview of sensortoolkit Evaluation Objects

• Conducting Analysis - The SensorEvaluation Class
  • Plotting Methods
  • Descriptive Summary Methods
  • Statistical Methods
• Creating Testing Reports - The PerformanceReport Class