Known pairs issues/questions about large-network-analysis-tools HOT 10 CLOSED

Hello @michaeldouble07.

For the date issue, here is what I believe is happening: When you choose a generic weekday as designated by the 1900 date, the solver actually considers this to be the next occurrence of this weekday. So when you chose Wednesday, it uses the next occurrence of a Wednesday, which in your case was June 8. The traffic evaluator has logic to figure out the probable traffic conditions for that date and time, and for a future date and time, it's based on historical trends. If you were using the current date and time, it would use live traffic. But anything a few days in the future will be using historical averages. You could also pick a random Wednesday in the past and likely get the same result. Unless there is live traffic data cached for that date, it will use the historical averages included in the network's traffic data. I see now that this behavior is poorly documented, and I'll see what we can do to improve that for the next release. (To be clear, this is core Network Analyst behavior, not something specific to my sample tool in this repo.)

Let me know if you need more detailed info on that, and I'll get one of my colleagues who knows more about this than me to comment.

Regarding curb approach: Hmm, I see that the way I've implemented tool does not transfer CurbApproach values from the input table automatically. I will need to update the tool to accommodate this. I don't anticipate this being a very difficult update. What is your timeline? When do you need this?

from large-network-analysis-tools.

Hello @mmorang

I need the curb approach feature, by the end of the week or beginning of next week. One thing I would like to clarify is that if I use the known pairs tool I should get the same results a as the large is matrix, is this correct? I also had a bit of trouble when I created the feature class and tried to do point calculations on the fly when adding to the feature class, it did not work with the known pairs tool, but it works for the large of matrix tool, so for the known pairs tool I had to make a table and then use table to point, and that worked for the known pairs tool. Could you please explain a valid config to create a feature class that would work with the known pairs tool.

from large-network-analysis-tools.

Okay, I'll see what I can do about the CurbApproach problem. Hopefully it will be really simple.

The known pairs and large OD tools should be similar, although it's possible the results might be slightly different in some edge cases. The known pairs tool is actually using the Route solver under the hood and not OD Cost Matrix. That's also why you might see different fields in the output table.

For the second half of your question, I don't understand what you mean by "when I created the feature class and tried to do point calculations on the fly when adding to the feature class, it did not work with the known pairs tool, but it works for the large of matrix tool". Could you clarify what you're trying to do and what error or bad results you got?

from large-network-analysis-tools.

Hi again. I've updated the code in this branch: https://github.com/Esri/large-network-analysis-tools/tree/transfer_input_fields

From my preliminary testing, this seems to be working to transfer valid input fields from https://pro.arcgis.com/en/pro-app/latest/arcpy/network-analyst/route-input-data-types.htm#ESRI_SECTION1_D64F7E7B5196448B9B64CDD603BF8207 into the analysis. So if your Origins and/or Destinations input has a valid field called CurbApproach populated with valid values, it will be used.

I would like to do some further testing before I merge this, but I think it is working sufficiently right now for you to get your project done. The main issue at the moment is that it might throw errors if the input includes fields with matching names but the wrong data types or populated with invalid values.

Please let me know if this works for your needs.

from large-network-analysis-tools.

What I was referring to before, when I made the feature class in python, I use python take data from a sql server and put it into a feature class for this tool. When I made the feature class like this arcpy.management.CreateFeatureclass( r"D:\Arcgis\MyProject9\MyProject9.gdb", "Destinations", "POINT", "Template", "DISABLED", "DISABLED", 'GEOGCS["GCS_WGS_1984",DATUM["D_WGS_1984",SPHEROID["WGS_1984",6378137.0,298.257223563]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]];-400 -400 11258999068426.2;-100000 10000;-100000 10000;8.98315284119521E-09;0.0002;0.001;IsHighPrecision', '', 0, 0, 0, '') with arcpy.da.InsertCursor(r"D:\Arcgis\MyProject9\MyProject9.gdb\Destinations", ["SHAPE@XY","CurbApproach"]) as incursor: for row in rows: incursor.insertRow([(row[1], row[0]),1]) del incursor

The known pairs tool does not like it with this feature class configuration. I know I must be doing something wrong, but I do not know what the known pairs feature class is looking for. So what I did to get around this is by making a table in Arcgis using python, then doing the table to point arcgis function. This is what I was trying to convey.

Okay, I'll see what I can do about the CurbApproach problem. Hopefully it will be really simple.

The known pairs and large OD tools should be similar, although it's possible the results might be slightly different in some edge cases. The known pairs tool is actually using the Route solver under the hood and not OD Cost Matrix. That's also why you might see different fields in the output table.

For the second half of your question, I don't understand what you mean by "when I created the feature class and tried to do point calculations on the fly when adding to the feature class, it did not work with the known pairs tool, but it works for the large of matrix tool". Could you clarify what you're trying to do and what error or bad results you got?

from large-network-analysis-tools.

That code looks generally fine. Did the resulting feature class look okay? All the points displayed in the map in the right places and it had the desired fields?

When you say "The known pairs tool does not like it with this feature class configuration," what do you mean? Did the tool throw an error message, and what was the message? I don't have enough information to diagnose that part of the problem.

from large-network-analysis-tools.

Hello @michaeldouble07. Were you able to successfully use your CurbApproach fields with my updated tool? I'd like to eventually merge this branch into the master branch, but I wanted to make sure it had worked correctly for you before I proceed.

from large-network-analysis-tools.

I haven't heard back from you, but I wanted to close out this project, so I went ahead and merged my changes. If you have CurbApproach in your input data, this should be used in the analysis successfully now.

If you're still having trouble with the other problem, please let me know. I'm happy to help. If I haven't heard back from you after some period of time, I'll close this issue as resolved.

from large-network-analysis-tools.

from large-network-analysis-tools.

I'm going to assume that you resolved your issue and that my updates to the tool to transfer usable input fields worked, so I'll close this issue. Feel free to reopen if you're still having trouble.

from large-network-analysis-tools.