For further context to the situation, as I explained previously, only 24 hours of data is kept in memory. This is done by fetching the most recent 24 hours of data initially, then splicing by the length of the new data received each time from the start of the dictionary – essentially keeping a fixed length dictionary of the points.
I have also implemented a method of changing the amount of data displayed on the graph through variables. If the limit were set to a minute, when updating the graph, the script would iterate through the stored 24 hours of data backwards until it find an item before the limit, at this point the data points for the graph will be set to a slice of the stored 24 hours of data from the found index to the last item.
I wanted to explain this in case the points on the graph were being mixed up somehow as a result of the shifting of the data points for the limits.
