I get a lot of value from Autodialer, but I donāt get a lot of value from having an accurate depiction of my yellow brick road 2 years ago. Can I safely expunge some of the old rate changes somehow?
The slowness is related to the graph itself, right? (And not downloading the json with lots of road changes or datapoints)
Well maybe you could keep the data while speeding up the page load by adjusting the graphās displayed date range (tmin and tmax). These are under your goal / Settings / Graph Settings / as X-min and X-max.
To clarify a bit further on Dannyās answer for anyone who isnāt super-experienced with the graph editor, you can delete quite a lot of the old graph rows with no trouble, as long as you make sure thereās an anchor point now to prevent your line changing place.
It actually didnāt work quite right at first, probably because I shouldāve set it not to todayās date, but to yesterdayās date, but the important thing to note is that you can see how it āanchorsā the line so that all the changes Iām making in erasing old rates donāt impact where the line is now. It shouldāve been yesterdayās date and not todayās so that it would keep things the same for the last deadline, and that wouldāve worked perfectly the whole time.
Make sure to click into the value column for that anchor row, like I did ā by default the rate is active and the value is auto-calculated, but you need the value to stay exactly what it is.
If you do want to make any other adjustments, like add an extra anchor and drag it around so that past datapoints arenāt in the red, you probably want to uncheck the button for āpropagate changes forwardā.
No changes are saved until you actually press save, so if you do something you donāt like, you can just hit undo.
But the trick mentioned about just changing the x-min would also work for you (though not for support if we ever needed to look at the graph in question ā for support things would still load slowly as we load in your whole graph matrix in our admin panel, even if itās hidden by the x-min setting).
Makes me think maybe this should be a feature of the autodialer to clean up after itself. On the other hand, I donāt know. Rewriting history feels bad.
I changed the graph to start two weeks ago and it didnāt seem to help. I never profiled it, but other parts of the page are set to scale with the number of rate changes too, like all the rate change text boxes on the settings tab.
I use the autodialer for many of my goals and have for a long time, so Iām curious why I havenāt noticed this problem on any of mine yet. Like given a comparable history of using the autodialer, shouldnāt it affect all goals perf equally?
I think it is not specific to the autodialer but comments have been made that the current implementation does not scale well^1 with many changes to the road matrix.
Probably the current implementation (beebrain) is such that it walks the entire datapoint and roadpoint (road matrix) lists validating datapoints, calculating rates, values, dot colors, and so on and currently must do so from the beginning til ānowā to create/calculate the graph (and stats).
Thus my curiosity about cropping the displayed graph range (tmin) does not carry the benefit I was thinking it might.
I usually use Safari, where I noticed the whole tab would freeze when opening my most heavily auto-dialed goal. Today, I opened that goal with Chrome and noticed a few differences:
The tab became interactive immediately after the page loaded.
The tab froze only momentarily when I opened the settings tab.
Chrome allowed me me to scroll the settings tab even before it had finished rendering all the text boxes. There was a big, empty space where the text boxes would eventually render, and they popped into place a few seconds later.
So it appears that both Chrome and Safari struggle with the page, but Chrome is better at doing work in the background.
My most successful was my exercise minutes goal. I knew that I wanted to get to 22 AZM per day, but I wasnāt anywhere near there when I started. The autodialer let me beemind a rate of improvement.
@alltom Are there a lot of duplicate rows in the goal matrix, like rows that could be deleted and not impact the graph even slightly? Also when was the goal created?
I archived it, and I guess I canāt see them any more. But the history was all fractional changes, for the most part IIRC. Iām sure that most of them had little impact.