6. Route segments with transportation data - ScandinavianSection-UCLA/hGIS_ETK GitHub Wiki
The route segments do not have any original attributes other than “absolute” and “abs_seq”- However, the means of transportation Evald Tang Kristensen used on the route segment is more relevant as an attribute of the route segments than of the stops. “abs_seq” is linking the route segment to the stop he is traveling to and enables the transportation data to be “joined” to the route shapefile.
- Right-click on the aggregate route shapefile > “Joins and Relates”
6.1 Complementing the transportation data from MO
The means of transportation originally extracted from the MO collections are ferry, train, horse wagon, walk and bike (on one occasion!). 1370/3913 route segments have this data, the rest of the fields of transportation attribute are empty. We have chosen to make sophisticated “best guesses” for these empty field.
6.1.1 Train
From a previous project, we have a train station point shapefile with the year the station opened and closed as attributes. We can intersect the route segment shapefile with this layer to identify route segments that passes through two or more stations that were opened during the year of the route segment was travelled on.
-
Select all the rows with missing transportation data in the route shapefile. Export the sub-set by right-clicking on the layer > “Export”.
-
Create a buffered train station layer: “Geoprocessing” > “Buffer”.
- Create a spatial join between the route layer and the buffered train station layer. It is no regular spatial join (“join by location”). Rather it is a tool called “Spatial Join” located in the Analysis toolbox.
- Each one route segment will have a record in the shapefile that is the output in step 3 as well as an additional record for each station that it is intersecting. For route segment with the unique identifier “abs_seq_2” “2002” there are two intersected stations:
- All the data from the station layer, as well as the route layer, is kept in the common attribute table of the output shapefile from step 3. This means the same record (route segment) has both the opening and closing date of the intersecting train station as well as the date the travel took place. “Select by attribute” allows us to choose only the records where the travel took place after the station opened.
Note:“Year” is the year Tang Kristensen traveled on the route segment. “Year_1” is the year the train station opened. No stations had time to close before Tang Kristensen was done with his collecting. For us, 3694/4576 places or train station intersections are confirmed to be valid in the attribute selection.
- In the attribute table of the new shapefile it is possible to perform a “summary” on an attribute of a field. It leaves the attribute table the same, but generates a new output table where every attribute that are of the same kind have been merged into one record per unique attribute. We are interested in the “abs_seq” field. Right-click on the attribute field > “Summarize”
Note: for example, “abs_seq” 9 will get the attribute 3 in the “count” field of the newly generated summarized table. This means three open stations intersects with that unique route segment.
Note: output table.
- We want to limit the table to include only route segments that should be assigned train transportation. This means only route segments that intersect with two or more train stations. “Select by attribute” is appropriate: “Selection” tab > “Select by attribute”.
Note: 769/3920 undefined route segments have been assigned train as mode of transportation in our model.
Note: this is how it looks like visually. The 353 stations are included.
-
The train assigned segments now needs to be connected to the original route attribute table to be used to update the attribute table with the new input. This is easy because abs_seq connects the two tables and can be used to perform a join: right-click on original route aggregate shapefile > “Join and relates”.
-
Select the records that had a “match” in the join. These are the new train route segments. Once selected, right-click on the original transportation field and use the “field calculator” to give the records the attribute “Train (model-assigned)”. Without the “ ” the field calculator will not accept text as an input.
6.1.2 Under 20-km walk
We have made it a rule that all remaining unassigned route segments under 20 km is likely to have been walking routes. The length of each segment can be established by adding a new field and right-click on the heading > “Calculate Geometry”. All unassigned segments under 20 km are selected and after right-clicking on the heading of the transportation field, the “field calculator” can be used to assign “Walk (under-20-km-rule)” as the attribute for the selected records. This becomes the value for 1699/3913 records.
6.1.3 Horse-wagon
All remaining unassigned route segments are assigned “Horse wagon (model-based)”.
Part two With all the data organized in aggregates, it is possible to do several different analyses. It is possible to group the data by attribute value of the stop or route, such as place he lived, year of travel, month of travel etc. It is also possible to combine the data with other data sources. In the case of our report, fine grain population data has been considered and can give an idea of how well represented different parts of Denmark have been.