ARTICLES

Using Python to Control MapPoint, Part 2

Second of a two-part article by Richard Marsden on scripting MapPoint using Python. In this part Richard shows how to use Python to plot the “250 largest cities in California”.

In Part 1, I showed how Python could be used to control MapPoint. Python is a scripting language that is ideally suited to text processing. Hence it can be used as a replacement for MapPoint’s Data Import Wizard in situations that require some processing intelligence. As an example, we will use it to plot the “250 largest cities in California”.

Our source data is the US Census Bureau’s 2000 Gazetteer of Places. This file (places2k.txt) and format specification can be found here:

http://www.census.gov/geo/www/gazetteer/places2k.html

The data file is a text file, with one line per place. A place is typically a town or city, although it may also be a CDP (Census Designated Place) – ie. an unincorporated town. Each line follows a fixed column format, and includes information such as location (longitude, latitude), area, and population.

So let’s get started with some Python! First we import the required libraries and MapPoint COM information:

# Import the required libraries
import string
from win32com.client import constants, Dispatch
MAPPOINT = 'MapPoint.Application.NA.11'

Python supports object oriented programming. This lets us represent a city’s information in the form of a simple object with only a constructor and some data members. This constructor takes a line of text (read from the data file) and creates a city object.

# This is a helper function used by the CityInfo constructor
def removeSuffix( str, suffix):
    idx = string.rfind(str,suffix)
    if (idx>-1):
        return str[0:idx]
    else:
        return str


# This class stores and handles a city's information
class CityInfo:
    def __init__(self, sline):
        self.state = sline[0:2]
        self.city = string.strip(sline[9:73])
        self.population = string.atol(sline[73:82])
        self.housing = string.atoi(sline[82:91])
        self.land_area = string.atof(sline[119:131])
        self.water_area = string.atof(sline[131:143])
        self.latitude = string.atof(sline[143:153])
        self.longitude = string.atof(sline[153:164])
        # trim the extra city/town/CDP/village off the end
        self.city = removeSuffix(self.city, " (balance)")
        self.city = removeSuffix(self.city, " city")
        self.city = removeSuffix(self.city, " CDP")
        self.city = removeSuffix(self.city, " town")
        self.city = removeSuffix(self.city, " borough")
        self.city = removeSuffix(self.city, " village")
        self.city = removeSuffix(self.city, " comunidad")
        self.city = removeSuffix(self.city, " municipality")
        self.city = removeSuffix(self.city, " zona urbana")

Note that Python’s tabulation is important. If a statement leads into a block, it ends with a ‘:’ and the members of the block are tabbed. So in the above code, the class CityInfo has one method, the __init__ method. This is the constructor self refers to the object being created, and sline is a string parameter holding the text line that is to be used. Most of the lines should be self-explanatory. The strip, atoi, and atof functions are a member of the string library. These are used to extract the string and numeric data from sline, before saving it into the relevant CityInfo members.

The city name needs a bit more work. Unlike the fixed 2 character State code, city names can have a variable length and are hence padded out with space characters. The call to strip removes these spaces. The US Census Bureau also appends a city type label to the end of each city name. Examples include “city”, “town”, and “village”. These are always l ower case and are removed by the removeSuffix() function, which is defined at the top of the above code snippet. This function simply looks for a particular suffix and removes it if it is present.

We will store the cities in arrays, but we also need to sort them. The array’s sort method can take custom comparison functions. We could write some quite complex comparison functions that sort on multiple parameters or calculate parameters to compare, but for now we only need to sort by population:

# comparison function to compare according to population
# Note that this is a reverse comparison: returns greatest first
# used for sorting
def sortByPopulation( a, b):
    return cmp( b.population, a.population )

Now we can start with the main code. First, we must read all the city data. These are stored as CityInfo objects in a dictionary called citiesByState.

######################################
# Main Code

citiesByState = {}
print "Example US Census Program: Reading data..."

# You will probably need to change the following path
cFile = open('c:\\Projects\\census\\places2k.txt', 'r')
city_line = 'empty'

city_line = cFile.readline()
while len(city_line) > 0 :
    thisCity = CityInfo( city_line )

    if (not citiesByState.has_key(thisCity.state) ):
         citiesByState[thisCity.state] = []
    citiesByState[thisCity.state].append(thisCity)

    # fetch next line (and loop)
    city_line = cFile.readline()

cFile.close()
print "US census data has been read"

This code simply loops over each line, reading it in and creating a CityInfo object with it (referenced by the variable thisCity). These are stored in the citiesByState dictionary. Python Dictionaries work a bit like STL maps or Visual Basic collections. They store collections of objects indexed by a key. In this case, we use the State as key, and for each key we store a list of CityInfo objects. This allows us to quickly find the list of cities for a particular State. Python lists work in a similar way as dynamic arrays.

After reading all these cities, we can perform a variety of actions – for example extract cities with specific population densities. However, here we simply want the 250 largest cities in California. So first, we get a reference to the list of California cities, and then sort them by population:

print "Sorting the California cities by population..."
myCities = citiesByState[ "CA" ]
myCities.sort(sortByPopulation)

Now that we have the cities sorted in order of population (largest first), we can extract the first 250 and plot them in MapPoint. Calling MapPoint in Python was covered in Part 1 of this article. Here is the city plotting code:

print "Starting MapPoint and creating pushpins..."

# start MapPoint, with a new map
myApp = Dispatch(MAPPOINT)
myApp.NewMap
myApp.Visible = 1
myMap = myApp.ActiveMap

# Loop over each city of these 250 cities, creating a pin for each
for iCity in range(250):
    thisCity = myCities[iCity]
    thisLoc = myMap.GetLocation( thisCity.latitude, thisCity.longitude)
    myPin = myMap.AddPushpin( thisLoc, thisCity.city )
    myPin.Symbol = 52

# Tidy Mappoint up, leaving MapPoint open for the user to see
myApp.UserControl = 1
myMap = 0
myApp = 0

print "Finished!"

The Python for loop is more flexible than the Visual Basic for loop, and is capable of iterating over a list of objects. Here we just want it to iterate over the range 1…250 – hence the call to range(250). Each pin is inserted at the city’s location with the city’s name.

This is what the resulting map looks like:

Click on the map to view the full screenshot

A similar script was combined with the Single State Mapper to produce the maps used on http://www.mileage-charts.com/cities/index.shtml.

Further Reading

"Learning Python", Mark Lutz & David Ascher, Publ. O’Reilly.

Prior to Winwaed, Richard worked as a software developer working on seismic processing algorithms for the oil exploration industry. He holds geology and geophysics degrees from the University of Cambridge (Churchill College), and the University of Durham; and an interdisciplinary MBA from the University of Dallas.

	MP2Kmag Internet