Pivot table instructions for capturing detection statistics (Internal)
How to use reports to find the 38 records at the same time…to see if it happens more often
You can extend this to run multiple companies and do the same thing, I was trying to be illustrative, not exhaustive.
Power
Note: I now can see average, max and min time between observations, per tag
It sucks pretty bad here in this example, but run some on blackbriar is it is dead 30 seconds every time almost It’s going to change based on if on host power (softap) or on battery….
What I do is use my router as a “control” and then I create another column called Detection % and I divide the Number in column d by the number in column d for control and then format to %
Should be 100% if host powered, above 100% if has a battery backup (TT-HBBLr1) as every 6 min the SSID changes, so you can get some “extra” detections w the reader, will be from 1 to 100% on battery depending on what the active/sleep time is.
Generally, we are doing 25ms on, 1225ms off for Battery backup and 25ms on and 1975 off for Battery and will get between 20 and 50% detections on the battery. You can see an example of this in the 2nd spreadsheet attached, see “sheet 2” worksheet, column 3 is this calculation. – this was a one-hour test, it is on dev, it sucks….as it is getting 40 out of possible 120 thirty-second timeslots in an hour……I can’t explain this. Look at 3rd spreadsheet this is from the product, it has 120 out of 120 possible observations, see “sheet 1” So I want to see the 64 observations from Cody's truck that are duplicated…double click on h8 and it creates a new spreadsheet w all the data (that is a sheet called duplicates from Cody's truck…it called it “sheet” I renamed it) Click on the pivot table and the box to the right comes up again so you can create/add more data Key, it happened on 3 of Cody’s 6 resources (observation at same exact time for an Asset) – this is NOT the same thing as the problem on the dev environment where we are seeing multiple observations in the portal and mobile app. How to use reports to find timing issues Create your report Add filters to your columns - , it’s under data Sort the tag id Troverlo Sort on column e, so all the tags together Customer Sort on column d Create a new column at the end – call it time delta Add this formula Troverlo =IF(E7=E8,F7-F8,0) Customer =IF(D6=D7,E6-E7,0) Copy all the way down Note, what this does is gives you a time delta between rows, but put a zero in it when the Tag ID changes Add new a column at the end – call it zeros Add this formula Troverlo =IF(U7=0, 1, 0) Customer =IF(T6=0, 1, 0) Copy all the way down Note, what this does is give you a count of all the rows w zero time delta in them Create a pivot table , it’s under insert Select checkbox at bottom It will create a new sheet Select Design, then subtotals, do not show subtotals Select design, report layout, show in tabular form Setup like this (it is on right side – drag the things from search into rows (3 of them) and drag tag id, time delta 3 times and count of zeros into values Click on first time delta, pick last option and select average Do same for next and pick max Do same for next and pick min Select three columns and set format to time xx:yy:zz Create new column after pivot – call it duplicate timestamps Add formula, =H4-1 Copy down You doing this because the last row of every tag will be zero, so you want to take it out of the calculation…see 6a above Now you have a pivot tab! After you copy over the config from testing guide sheet Column g - active cycles per test period =(C$29*60*60)/(K5/1000) c$29==# of hours in the test K5==sleep time in ms Column h == % of detection based on "active cycles in a test period =IF(G4>D$29, C4/D$29, C4/G4) D$29==# of sessions in test (so if 6 hours, 6*60*60/2 if 30 second cycles) C4 == count of detections, G4 == active cycles per test