Today I recieved a dataset of road lines which I plan to use for routing purposes. To my horror I discovered that the road lines where not captured according to any logical standard (some road lines are split midway, some roadlines go through intersections, some don't, some roadlines are split at intersection and some are not etc. etc).
I should also mention at this point that the data table do not contain any information which can be used to join these fragmented road lines. So what I probaly need to do is iterate through the dataset and join all lines which has the same end and start node. This will however also cause t junctions that end in the line to become part of this join.
a. Does anybody know of a automated way to join all lines that have the same start and end node. b. T junctions end in some of the intersection so they will inevitably be included in the above mentioned join, does anybody know of a way to counter this.
Any help will grately be appreciated. I am new to the world of routing (which sofar has proved to be depressing) and I do not want to spend hours manually editing the dataset when there is a easier way to do it.
Gert, [INDENT]As a first step you need to prepare your data so that it can be used to build a Network Dataset. Once you have a Network Dataset, you can perform routing or any other network analysis.[/INDENT] [INDENT]Your have to carefully analyze the street data and then do the necessary edits to ensure that the street centerlines have correct connectivity information that is required for building network datasets. While there is no single magic tool that will take your data and correct all errors in it, you can do this in a semi-automated way. The first step is to ensure that the lines connect end to end. If this is not the caase, you can run the Integrate geoprocessing tool. This will combine all the line ends if they are within the specified tolerance. You might have to try different values to determine the tolerance that works best for your data.[/INDENT] [INDENT]After integrate, you should build a topology from your street centerlines and set the necessary rules. This should further clean up your data.[/INDENT] [INDENT]Once the topology is correct, you can create a new network dataset. There are various options that you can use when creating a network dataset. The help topic has additional pointers to related topics.[/INDENT] [INDENT]Note even after you have edited the data to have the necessary connectivity, you will still need some attributes on streets centerlines such as Oneway street information as well as speed limits to model travel time. This is required to set up network attributes that are used during analysis.[/INDENT] [INDENT]If all this sounds too daunting, you can start with a free network dataset provided on Esri Data and Maps DVD that comes with ArcGIS. You can find it within the Streetmap_na\data folder on the DVD. This is a prebuilt readonly network dataset for whole of US and Canada. You can add the "Streets" network dataset to ArcMap from within this folder, enable the Network Analyst Extension and start performing network analysis. The data is about 5 years old, but it provides a good starting point.[/INDENT] [INDENT]I will also recommend to have a look at the technical workshop about creating network datasets presented at the Esri User conference. This should provide you with concepts necessary to build a good network dataset. [/INDENT]
The following tools might be helpful as well (requiring ArcInfo License): 1. Feature To Line tool (Data Management toolbox - Features toolset) breaks lines at intersections. 2. Dissolve with the UNSPLIT option and the new tool, Unsplit Line, in ArcGIS 10 remove pseudo nodes (lines with end-to-end connections, not at T- or cross-intersections).