Case08A-Spatial Distribution and Clusters of Place Names in Yunnan

Chapter 8 Spatial Statistics and Applications

Spatial statistics analyzes the pattern, process and relationship in spatial (geographic) data. This chapter implements some fundamental tasks in spatial statistics: measuring geographic distributions, spatial cluster analysis, and spatial regression models. Section 8.1 uses a case study on place names in Yunnan, China, to illustrate the implementations of various centrographic measures and point-based spatial clustering. Section 8.2 employs a case study to demonstrate the methods for detecting global and local colocation patterns of two types of points (here crimes and facili ties). Section 8.3 applies spatial cluster analysis and spatial regression in a case study of homicide patterns in Chicago.

Case Study 8A: Spatial Distribution and Clusters of Place Names in Yunnan

This project is based on a study reported in Wang, Zhang, et al. (2014). Similar to the case study 3A on Zhuang place names in Guangxi of China in Chapter 3, this case study also examines place names and shifts the study area to Guangxi’s neighboring province, Yunnan. The place names in Guangxi are largely binary and grouped into Zhuang and Han. Yunnan has many ethnic minorities and thus a mix of multiethnic toponyms. Analysis of multiple types of point data presents unique challenges and affords us an opportunity to showcase the values of spatial statistics in historical and cultural geography. The case study has two parts: Part 1 implements the centrographic measures to describe the spatial patterns of toponyms, and Part 2 identifies whether concentrations of one type of toponyms in some areas are random or statistically significant.

Among a total population of 44.5 million in Yunnan, two-thirds are the Chinese majority Han and one-third are ethnic minority groups. More than a quarter of current toponyms in Yunnan are derived from ethnic minority languages. The multiethnic toponyms in Yunnan are grouped into three language branches: Zang–Mian (Tibeto–Burman), Zhuang–Dong (Kam–ai), and Han, all of which belong to the large Sino-Tibetan language family.

Data for the case study is organized in geodatabase YN.gpkg under the folder Yunnan that includes
1. Three reference layers: Yunnan_pro (boundary of Yunnan Province), Prefectures1994 (Prefectures in Yunnan), Majorcity (Kunming and Dali).
2. Layer PlaceNames with (1) a field Large_grou identifying place names of three language branches (han, zang-mian, and zhuang-dong) and (2) two fields PointX and PointY representing the XY coordinates.
3. Four layers for Han place names corresponding to four different eras: Han_PreYuan (before 1279), Han_Yuan (1279–1368), Han_Ming (1368–1644), Han_Qing_postQ (after 1644).

The R files for Spatial Scan (Kulldorff scan in SatScan) are in the same folder: SatScanR-1.r, SatScanR-2.r and SatScanR-3.r.

Computational Methods and GIS Applications in Social Science - KNIME Lab Manual
Lingbo Liu, Fahui Wang

External resources

• Github for Geospatial Analytics
• Github for Workbook Issue Report