Cycle Length & Vertical Scale question

[jkosky]

I’ve got a question that I wasn’t able to find in the User’s guide or perhaps didn’t understand it.
As you can see in the screenshot attached that the cycle length says “90s” but in reality the cycle lengths are 120s three intersections and 60 for two of them. I would like to change the diagram to be able to see 120s instead of 90s without having it adjust anything else but the vertical scale. Is it possible?

[bullock]

Open the Arterial Diagram Parameters dialog.
Under the Diagram tab, you may enter the scales, both vertical and horizontal.
For the vertical (time) scale, you may either set a specific value or select a number of (network wide) cycles to show on the diagram. In the latter case, as you change the cycle length, the scale gets adjusted automatically to maintain the specified number of cycles visible.

[jkosky]

I found that, yes, it will allow me to see more cycles, but I would like the number 90s to show 120s and have the scale match it.

[bullock]

Sorry, I'd misunderstood.

Unfortunately, to show 120s as the network-wide cycle length might involve quite a few more steps, depending on whether you need the individual intersections to retain their current cycle length or whether you want to retain their harmonic relationship to the network-wide cycle length.

You may set the network-wide cycle length in the Network Parameters dialog. An easy way to navigate there is to single click on the cycle length near the upper left corner (the part you highlighted in yellow).

Once you change that to 120s, the diagram(s) will indeed show 120s in the upper left, and the vertical scale, if it's not locked, will automatically adjust as needed so three full cycles still show. All of that is probably as desired.

Alas, a few things you may not desire will also happen:

1. The intersections (all those not specifically set to an uncoordinated cycle length) will change to a new cycle length such that the ratio (the harmonic relationship) of their local cycle length to the network-wide cycle length is preserved, and
2. The splits will all change according to the "Split Type" -- essentially the scaling rule -- which may be specified individually for each movement, and may be set to Freely Scaled, Fixed Seconds, or Fixed Percent.

Assuming you want the intersections to all retain their current cycle lengths, you'll need to step through each intersection and change its Adjusted Cycle Length (specifying the harmonic relationship to the network-wide cycle length) such that the cycle length is back to where it should be. Fortunately, as you do this, the splits will all scale back, using the same scaling rule as before but in reverse, reverting to their original values.

[jkosky]

I see. It probably isn't worth my time to make all of those ammendments. The current view still works though so I'll probably just keep it the way it is. Thanks for the reply. =)

[bullock]

This might be your lucky day.

As it so happens, after thinking about this further, I added a new option to the developmental version of the software -- the version that is becoming v11.0. Now, when you change the network-wide cycle length to a harmonically related one, the software offers to either

• Scale the local cycle lengths. Preserve their current, harmonic relationships to the network-wide cycle length, so 90s → 120s, 60s → 80s, etc., or
• Preserve the local cycle lengths. Make compensating adjustments in the local, harmonic relationships to the network-wide cycle length, preserving the local cycle lengths, so 90s → 90s, 60s → 60s, etc.

If you'd like, you can email your *.Dgm file to me, and I'll use this new option in the developmental version to change the network wide cycle length from 90s → 120s, while preserving all the local cycle lengths (90s → 90s, 60s → 60s, etc.). I'll email back to you the modified file.

[jkosky]

That is great! And it would be a nifty presentation feature. I'll shoot those .dgm files over to you. Thanks Greg!