For the details of how-to calculate each approach, see session 3 below.
My original plan was to truly present all data upto last minutes, but that comes with a caveat of more loading on the server. A longer loading time is a practical consideration. A improvised plan is to serve static correlation charts of the current data of each month. Each chart is calculated at the mid-night of each day with the data of the current month upto the previous day.
Below are six graphs for 3 pairs with Dylos DC1100 as the "reference". And -ad indicted the concentration of PM2.5 was adjusted with a MetOne 1020 FEM monitor. The adjusting coefficient was reported in another study linked here.
Here are some graphs of correlating between HPMA115S0, PMS7003 and SDS011 with Dylos DC1100 Pro.