My final project mainly focuses on how to achieve dataset available for network analysis. Taking advantage of maps offered by OpenStreetMap (http://planet.openstreetmap.org/), specific New York City map extracted by BBBike (http://download.bbbike.org/osm/) and great map importing tool OSM2NetworkDataset (http://52north.org/communities/geoprocessing/osm2nds/). After successfully built a network dataset for New York County (clipped the geo-database from New York City), I geocoded top 8 museums in Manhattan and used tools of network analysis as follows:
Routing (First stop is set to be Grand Central Station and last stop is American Museum of Natural History, optimizing route by reordering stops of museums)
Closing facilities (3 nearest subway stations to American Museum of Natural History)
OD Matrix (From Stations to Museums)
Service Area (Within 3 minutes drive to museums)
Another analysis I made is to find the most appropriate location for future construction of museums, basing on the parameters as follows:
Median age which I assumed the lower median age is, the higher demand for museums nearby
Number of people employed in art industry in each census tract which I assumed the more people employed in art industry, the higher demand for museums nearby
Distance from existing museums which I assumed the farer from existing museums, the higher demand for museums nearby.
I projected the original shapefile from WGS to UTM Zone 18N
Turned the data in the census tract into raster
Calculated the Euclidean distance of each tract to the existing museums
Reclassified these three parameters and used raster calculator to sum up them to get an index of priority for building future museums
Originally, I intended to calculate the most optimized route to every national park in the United States. However, I found that it is hard to achieve the speed data and route restriction data which are necessary for network analysis.
Then I achieved the network dataset for my favorite city of Beijing and wished to make hotspot analysis as well. Unfortunately it is almost impossible to find or make address locater for the city and neither is dataset of census. I only located my home and my university in the map and found one optimal route between them.
All information about the road is from database of OpenStreetMap. I googled the addresses of museums and geocoded them in ArcGIS. The shapefile is NY County 2000. The attribute data of median income is from 2010 Census Summary File 1. Number of people employed in art industry is from ACS_11_5YR_DP03.
Most difficulties I faced are due to little knowledge of java command line which is essential for any tool to deal with maps. Since the time of achieving a result from java programs is also very considerable, I waited for hours and hours and possibly achieved an error leading to the failure of the program. Thanks to google and related forum in Arcgis website, I finally managed to make my final project presented here.
Thank you Yoh! Thank you Zhongbo! I am pretty much sure I would like to take the advanced course in the spring quarter (sadly it is also my last quarter here).