ETL - Calibration analysis
Sources and plots for some of the analysis related to SE and AirSensEUR Calibration.
- josene - plots for ANN Calibration of Intemo Jose sensors
- airsenseur - regression plots for calibrated and reference data for AirSensEUR
airsenseurplots.py draws various plots from data stored in InfluxDB.
Both calibrated and RIVM reference data are stored there. For each gas both R-squared ('R2') and the slope ('m') is
depicted. The Python
scipi.stats.linregress(x, y) function is used.
Two periods and locations are taken:
- sept 9 to okt 9 2018 at RIVM Breukelen Snelweg (A2 highway) site where all 5 AirSensEURs were placed
- dec 25 2018 to january 24 2019 at RIVM Nijmegen Ruyterstraat where ASE_NL_01 was placed
Especially for ASE_NL_01 the calibration can be extra evaluated as this station was present at two RIVM locations. The result is shown in a combined plots for ASE_NL_01. Upper row is in Breukelen, lower in Nijmegen (only Nijmegen has RIVM CO via SOS)
For all ASEs, especially NO2 prediction is quite good: all slopes ('m') near 1 and R2 all above 0.9 (except ASE_NL_02).