Being Clever with Data

2 April 2000
Yesterday I found that the Geographical Names database was accessible by FTP, so that I could download the entirety of a state's entries. The files are pretty big, too big for Word and Excel to handle, so extracting something like 'Hollow' is not feasible that way. It occurred to me that the UNIX grep command might be just the thing, and so it proved to be. Here are the steps:
  1. ftp the desired file from USGS (using the Mac) and FETCH it to an appropriate directory on Liberty
  2. gunzip the file
  3. to extract 'Hollow' what I typed was
    grep Hollow TN_deci > tnhol
  4. the resulting file can be opened in Word and cleaned up (the "," delimiters changed to ^t, and the extra "s removed)
  5. and then opened in Excel. The great advantage is that the coordinates are already in decimal degrees.
In Excel I removed the 'Hollow' records which are not coded as 'valley' (leaves 4517...), copied the resulting matrix, did a 'Paste Special...', choosing 'values', and then changed the format of the decimal degrees columns to a uniform 3 decimal places, and saved the product as a .dbf file (it's D:\esri\tnhol\tenvalhol2.dbf). Need also to add COLUMN LABELS as a new row.
I extracted the TN counties from the US county dataset (it's in the ESRI data, D:\esri\esridata\usa\counties.shp), then added the TN hollows data:
  1. Tables icon in the Project window
  2. "Add Table"
  3. View menu: "Add Event Theme"
  4. and hey presto...