8. Stop distribution - ScandinavianSection-UCLA/hGIS_ETK GitHub Wiki

8.1 Stops by residence at the time

Simply displaying the stops by which of the five different places he lived gives a chronological overview of his collecting. The order of the places he lived is Gjellerup (1866-1876), Faarup (1876-1884), Brandstrup (1884-1888), Hadsten Station (1888-1897), and Mølhom (1897-1929).

8.2 Stops by sogn

With multiple stops at the same location, it is easier to get the geographical distribution of the stops he made by counting number of stops in an area. In the case of this project, the stop data has been overlaid by the Danish sogn borders. However, first we need a stops shapefile where his origins of travel are not overrepresented:

  1. The first and the last stop of his travels needs to be removed from a new copy of the stops shapefile. To get a new shapefile, right-click on the stops shapefile > Export > give name and select location of new, to be reduced, stops shapefile.

  2. The “absolute” field in shapefile has the same value for all the stops in one fieldtrip. Summarizing that field and chose to get the minimum and maximum “abs_seq” value gives me the abs_seq value for the first and the last stop in each fieldtrip.

  1. The output table from step 2 can be joined back to the new stops shapefile by the abs_seq values. And in turn all the matching records can be selected and deleted. The generated stops shapefile will be used in all the following stop distribution analysis. The findings can be presented as numbers in a chart, showing the number of stops within every sogn, or visually with preferably a choropleth map:

  2. Right-clicking on parish boundary layer (sogn) > Joins and Relates > Join… > Choose the option to perform the join based on spatial location. The layer to be joined to the layer is the stops data.

  1. The number of points that fall inside will be a new attribute in an output shapefile, which visually looks like the original sogn shapefile.

  2. Right-click on the new sogn shapefile > Properties > Symbology > Quantities > Graduated Colors. Choose to base the colors on the attribute field representing the number of points that fall within the sogn.

8.2.1 Stops by sogn per km2

With the attribute of number of that he made in each sogn, the data can be normalized by calculating stops per km2. This is appropriate since the sogns are of different size.

  1. The number of stops needs to be divided by the area of the sogn in square kilometers. The area can be calculated in an added field: Right-click on sogn layer > Open attribute table > > Add field > Right-click on field heading of new field > Calculate Geometry.

  1. Add another field that will contain the calculated stops per square kilometer. Use the field calculator to divide the stops field with the newly generated area field: Right-click on the field heading > Calculate Geometry.

  2. Create a choropleth map with a graduated color scheme (see step 3, 8.2 Stops by sogn).

  1. A tool in the spatial analyst toolbar called “Point Density” can be used as an alternative to show the point distribution. The default settings look good.

8.2.2 Stop distribution and population

The stop distribution can also be considered in terms of how well represented the population is of Tang Kristensen visits. The population data layer was generated in 2014 and separate metadata and procedures is accessible upon request. In the map below the population data of 1890 is shown. We have separate data for the rural area and the market cities here combined.

  1. To combine the population data with the number of stops per people living in the population unit, we need to divide the number of stops that fall inside the population unit by the population number living within the boundaries.

  2. Right-click on the population layer > Joins and relates > Join by spatial location:

  1. Now the number of stops as well as the population data is attributes in two separate fields. Number of stops per people will be too small of a unit. Rather, add a new field > field calculator > new field = population/1000.

  2. Add another new field > field calculator > stop count field/new per 1000 population field. Now we can generate a map that shows number of stops Tang Kristensen made per 1000 people living within the population unit:

8.3 Repeat visits

The stop distribution in 8.2 does not show points that are multiple in the same location. In to better show this, multiple points in the same location should be shown as an attribute for the point. If it is an attribute, we can generate both maps where the size of the point correspond to the number of visits to the location as well as different types charts where the top places are visited are listed.

8.3.1Repeat visit point map

  1. With multiple places having the same name, we need to start by creating a unique identifier for each individual place location. Each point on a map has a geographic location represented by a latitude and longitude coordinate which can be shown as two separate attributes by adding a new field > calculate geometry > X coordinate of point. Do the same steps to get the Y coordinate.

  1. Add yet another field which will create a unique identifier for a location by combining the X- and Y coordinate fields. Field-calculator > new field = X + Y fields

  1. Summarize by the unique identifier to get a count of stops with the same location. The summarize function generates a table without geographical location. To get the data back into ArcMap as a shapefile with one point for each location, the longitude and latitude values are also included in the output summarized table:

  1. Add data > Add XY data to get it back into ArcMap.

  1. Table of the stops that are visited ten times or more:

8.3.2 Repeat visits by year

Over time, it is natural that he had to start come back to places that he had already been. However, he does this to a higher degree than probably necessary seeing that he gets to know places and people that he found productive and has motives to come back. There is a way to see how big percentage of visits every year were repeat visits. The total number of visits during a year will be compared to the number of repeat visits for the year and shown in percentage.

  1. With the point layer containing only one point for each unique location vitiated and a count for each location, the places visited more than once can be excluded. Select by attribute > “Count” field >= 2. Export the selection by right-clicking on the shapefile > “Export”.

  2. Excluded the times he visited the place first: summarize by the field with the combined latitude-longitude value (see 8.3 step 3) > choose to get the minimum abs_seq value in the exported table.

Note: the look of the summarize table.

  1. The exported table will contain the abs_seq value for all the first-time visits. The table can be joined back to the abs_seq attribute of the repeat visit layer: right-click on repeat visit layer > joins and relates > join data from the summarized table from step 2.

  2. All the successful joins are first visit records. There are 543 of these. Select the all and chose to delete them. Left is only stops that he is returning to.

  3. Use summarize on the reduced repeat visit shapefile from step 4. Summarize the year field to get a count of number of repeat visits he performed each year.

  4. Summarize the original stops shapefile (the one with no first and last stop – see 8.2) by the year field to get number of stops per year.

  5. Combine the two tables and divided the number of repeat visit with total visits in a new field. The percentage of repeat visits each year can then be shown in a graph:

8.3.3 Repeat visit leader chart

In the same way that a count of stops per year can be extracted using the summarize function on the combined latitude-longitude field (see 8.3.2), the summarize function can be used to get a leader chart of which places that was the most visited during different time periods of Tang Kristensen’s collecting. Repeat for each time-period:

  1. Use the function “select by attribute” on the “year” field to select data from the first time-period of interest, 1868-1878.

  2. Summarize the combined latitude-longitude field by right-clicking on the field heading > “Summarize” > Select “first” for the field name (should be the same name for all the records with the same combined latitude/longitude value). A “count” field will be part of the output table automatically.

  1. Export the output table from step 2 to excel to make a tables and graphs.

  2. Past all the data from the different time-periods into the same excel table. Sort and arrange as preferred.