By adding the ‘%’ in the systemmonitor.py you’ll see an graph:
Change that to:
‘load_15m’: [‘Average Load (15m)’, ‘%’, ‘mdi:memory’],
‘load_1m’: [‘Average Load (1m)’, ‘%’, ‘mdi:memory’],
‘load_5m’: [‘Average Load (5m)’, ‘%’, ‘mdi:memory’],
By adding the ‘%’ in the systemmonitor.py you’ll see an graph:
Change that to:
‘load_15m’: [‘Average Load (15m)’, ‘%’, ‘mdi:memory’],
‘load_1m’: [‘Average Load (1m)’, ‘%’, ‘mdi:memory’],
‘load_5m’: [‘Average Load (5m)’, ‘%’, ‘mdi:memory’],
The system load has no unit. Thus the code is correct. If you want to graph the load then take a template sensor.
Agree the system load should have no unit. But it seems like the lineChartDevice/timelineDevice decision should be configurable and not simply based on whether the unit_of_measurement is specified or not.