Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Make BetterBusBuffers work with the Network Analyst public transit data model #137

Closed
mmorang opened this issue May 20, 2019 · 24 comments
Closed

Make BetterBusBuffers work with the Network Analyst public transit data model #137

mmorang opened this issue May 20, 2019 · 24 comments
Labels
BetterBusBuffers enhancement

Comments

@mmorang
Copy link
Collaborator

mmorang commented May 20, 2019

Instead of converting GTFS into a sqlite database, rewrite BetterBusBuffers to use that tables in the Network Analyst public transit data model added in ArcGIS Pro 2.4.
@mmorang mmorang mentioned this issue May 20, 2019
Open
@d-wasserman
Copy link
Contributor

What are the accessibility trade off's of doing this? It is not a huge concern as these tend to be intermediate files, but will licenses be required?

@mmorang
Copy link
Collaborator Author

mmorang commented Nov 7, 2019

I had considered doing this as a first step to getting BetterBusBuffers into core ArcGIS Pro. The advantage is it could use the GTFS processing code that's been built out in Pro for doing GTFS validation, and then users wouldn't have to process their GTFS data in two different ways if they were also using the new Pro network dataset stuff. However, I haven't really gotten any requests for this, so I don't plan to do anything on this in the immediate future.

@d-wasserman
Copy link
Contributor

I would not call it a high priority ask, but I see the value.

@mmorang
Copy link
Collaborator Author

mmorang commented Nov 7, 2019

And what is that value to you? Anything beyond what I just said? This is important for me to understand.

@d-wasserman
Copy link
Contributor

I think the value to me is as the ArcGIS Pro GTFS Validation & Prep procedures improve, we don't have to have a split methodology to reference in terms of how the GTFS is (pre-) treated. It does seem somewhat disconnected. It seems like it has long term potential to help move the needle on other things I have complained about on the issues pages. =)

@mmorang
Copy link
Collaborator Author

mmorang commented Apr 28, 2020

Someone has asked for this functionality: https://community.esri.com/message/925240-better-bus-buffers-using-public-transit-enabled-network-dataset

@d-wasserman
Copy link
Contributor

Again, my licenses question is one that I want to understand, but I see the value for a streamlined workflows.
Regardless of the question, I put this still in a low priority camp. Is this nudging the priority for you?

@mmorang
Copy link
Collaborator Author

mmorang commented Apr 28, 2020

A little, but I'll have to brainstorm it a bit in the coming weeks. I'm not sure if I would even want to port the toolbox (or a subset of its tools) into core Pro or just update the downloadable version to work with the data model. And I have no idea how the licensing would work since you wrote one of the tools. It's still fairly low on my priority list, but it's higher now that somebody has asked for it.

I have never been happy with the procedure I use for generating the overlapping polygons. I've tried numerous times to make that work better to no avail. I would not feel comfortable making the polygon tools into core Pro tools unless I could do something better.

@d-wasserman
Copy link
Contributor

d-wasserman commented Apr 28, 2020 •
edited
Loading

Hi Melinda,
I gave that tool to the repo under a shared Apache 2.0 license. The attribution requirement is pretty minimal, and I would be fine with it in the code as it is now. I don't think the interface necessarily needs to change (though I would not mind a shout out in a press release or a tool tip).
At a more fundamental level, we would need to change the tools to work with the new data model. I am willing to help with that if you want some outside assistance.

Why not just high-jack the methodology used by the collapse polygon tools we iterated on together? You would give the user control over output resolution that way too. Not ideal I recognize, but it could be a solution.
You could also let the user create a sampling geography of their choosing.

@mmorang
Copy link
Collaborator Author

mmorang commented Apr 29, 2020

If I remember correctly, the methodology we worked out for Create Percent Access Polygons does a really nice job of getting a raw count of polygons overlapping each cell. For BetterBusBuffers, though, I need to avoid double-counting trips. If an area has access to two transit stops, but those stops are along the same line and consequently are served by the same trips, then it should count as 1 instead of 2. So, each service area polygon needs to maintain the stop_id it's associated with, and not only does it need to have a trip count, but I need to know the actual IDs of the trips so I can remove duplicates. It's very messy.

@d-wasserman
Copy link
Contributor

Ok, got it. I was thinking it is a sampling and trip counting problem, but maintaining the stop id of the original service areas sounds complicated. That information would be lost in a raster transform unless you allocated it back in some fashion (which sounds like a potentially larger headache).

@mmorang
Copy link
Collaborator Author

mmorang commented Jun 8, 2021

For anyone following this issue: BetterBusBuffers will be added to core ArcGIS Pro as the Calculate Transit Service Frequency tool in the 2.9 release, scheduled for late 2021 or early 2022. It will use as input the Public Transit Data Model tables and feature classes.

@mmorang
Copy link
Collaborator Author

mmorang commented Nov 11, 2021

This was completed, but not here in this repo. ArcGIS Pro 2.9 was released today, and it features the new Calculate Transit Service Frequency tool, which does similar things to BetterBusBuffers but uses as input data in the Public Transit Data Model format.

BetterBusBuffers is now deprecated.

@d-wasserman Thanks for all your input on this tool in the past. It's been awesome! (And you know where to find me if you have more ideas in the future.)

@mmorang mmorang closed this as completed Nov 11, 2021
@d-wasserman
Copy link
Contributor

All good things have their time. It was real fun to work on this project over the years with you. I am sure we will find another problem to abuse the ArcGIS Ecosystem with.

@paulogeo
Copy link

Hello guys. I know BBB is now deprecated, but i used it in my thesis and i would like to know if any of you guys know if there´s a "formula" to it, like all those formulae of potencial accessibility or cummulative opportunities...

Thank you all for the help.

@d-wasserman
Copy link
Contributor

Much of the tools live on in the Calculate Service Frequency Tools.

If you are asking if there are decay functions for access potential like those documented in the Transportation Access Manual, the answer is no. You likely derive a relatively easy to compute approximation using "median isochrones" and breaks that enable you to apply stepwise decay functions. However, the main purpose of these tools main purpose are to enable users to deal with the temporal variation of transit in a network analysis context.

@mmorang can confirm I am correct, but my understanding is these tools only do "flat" aggregations when calculating cumulative opportunities.

@paulogeo
Copy link

paulogeo commented Jan 14, 2024 via email

@d-wasserman
Copy link
Contributor

Ah you mean a citation.

Morang. M.; Wasserman D. (2015). better-bus-buffers. Esri. Retrieved from https://github.com/Esri/public-transit-tools.

@mmorang might have a second thought to add on this.

@paulogeo
Copy link

paulogeo commented Jan 14, 2024 via email

@d-wasserman
Copy link
Contributor

I think I need more information about the type of analysis your thesis did then. Again, many of these tools are simply about temporal normalization of transit schedules in physical space. If you did an access analysis, many of these tools operate to provide effectively "percentile" representations of access, so the mathematical formulations I guess would be similar to a percent of access times?

@paulogeo
Copy link

paulogeo commented Jan 14, 2024 via email

@d-wasserman
Copy link
Contributor

I go back to my decay function note. These tools only do flat access aggregations by default. Melinda's percent of access opportunities tool does not do any decay operations. I think it is a naive sum of connected opportunities within a percentage of time points.

@paulogeo
Copy link

paulogeo commented Jan 15, 2024 via email

@mmorang
Copy link
Collaborator Author

mmorang commented Jan 16, 2024

Just catching up on this thread. @paulogeo, it might help if you told us which of the BetterBusBuffers tools you used in your thesis. They all do similar things, but, as David mentioned, they aggregate the results in different ways.

In general, the tools simply provide counts of transit service. They simply check the schedules to determine which transit trips visit stops during the provided time window and summarize that service. The part that depends on the tool is how those counts are aggregated. For example, Count Trips at Points calculates an OD Cost Matrix to identify which stops are within the designated cutoff time/distance of each point of interest and combines the data from all those stops. It doesn't use any kind of decay function to weight the stops by distance. It just counts them all equally.

Full documentation for all the BetterBusBuffers tools is at https://github.com/Esri/public-transit-tools/blob/master/deprecated-tools/better-bus-buffers/UsersGuide.md.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
BetterBusBuffers enhancement
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@d-wasserman @mmorang @paulogeo