In [176]:

    

import mpl_toolkits
import mpl_toolkits.basemap

#
# extract tornado 2010 data
#
with open('torn.csv', 'r') as torntxt:
    lines = torntxt.readlines()

#
# extract lists of latitudes, longitudes, and scales
#
tornTuple = [line.split(',')  for line  in lines]
lats      = [float(tuple[15]) for tuple in tornTuple]
lons      = [float(tuple[16]) for tuple in tornTuple]
scales    = [int(tuple[10])   for tuple in tornTuple]

#
# quickly sanity-check the data to make sure it is sensible
#
figsize(15, 3)

subplot(141)
title("Distribution of Latitudes");
hist(lats, bins=20);

subplot(142)
title("Distribution of Longitudes");
hist(lons, bins=20);

subplot(143)
title("Distribution of Tornado Intensity");
hist(scales, bins=5);


    

In [140]:

    

#
# specify the map boundaries and projection type
#
map_func = mpl_toolkits.basemap.Basemap(llcrnrlon= -120,
                                        llcrnrlat=25,
                                        urcrnrlon=-50,
                                        urcrnrlat=50,
                                        projection="tmerc",
                                        lon_0 = -95,
                                        lat_0 = 35,
                                        resolution = "l")

figsize(8, 8)

#
# draw some features of the map (for more fill options refer to the Basemap Resources in the HW4 spec).
#
map_func.drawcoastlines(color = "gray")
map_func.drawcountries(color = "gray")
map_func.drawstates(color = "gray")
map_func.fillcontinents(color = "beige")
map_func.drawmapboundary();


    

In [178]:

    

figsize(8, 8)

#
# plot in a single color with slight transparency
#
title("Tornado Occurances")

map_func.drawcoastlines(color = "gray")
map_func.drawcountries(color = "gray")
map_func.drawstates(color = "gray")
map_func.fillcontinents(color = "beige")
map_func.drawmapboundary();

map_func.plot(x, y, "o", alpha=0.25);


    

In [179]:

    

#
# decide how to color each point
#
color_map   = matplotlib.cm.get_cmap("Blues")
norm_scales = [float(scale)/5          for scale      in scales]
my_colors   = [color_map(1-norm_scale) for norm_scale in norm_scales]
my_sizes    = [norm_scale*30           for norm_scale in norm_scales]

figsize(8, 16)

subplot(311);

title("Tornado intensity encoded as marker size")

map_func.drawcoastlines(color = "gray")
map_func.drawcountries(color = "gray")
map_func.drawstates(color = "gray")
map_func.fillcontinents(color = "beige")
map_func.drawmapboundary();

for i in range(len(tornTuple)):
    map_func.plot(lons[i], lats[i], 'o', color="blue", alpha=0.25, markersize=my_sizes[i], latlon=True)
    
subplot(312);

title("Tornado intensity encoded as color")

map_func.drawcoastlines(color = "gray")
map_func.drawcountries(color = "gray")
map_func.drawstates(color = "gray")
map_func.fillcontinents(color = "beige")
map_func.drawmapboundary();

for i in range(len(tornTuple)):
    map_func.plot(lons[i], lats[i], 'o', c=my_colors[i], latlon=True)

subplot(313);

title("Tornado intensity encoded as size and color")

map_func.drawcoastlines(color = "gray")
map_func.drawcountries(color = "gray")
map_func.drawstates(color = "gray")
map_func.fillcontinents(color = "beige")
map_func.drawmapboundary();

for i in range(len(tornTuple)):
    map_func.plot(lons[i], lats[i], 'o', markersize=my_sizes[i], c=my_colors[i], latlon=True)


    

In [180]:

    

minx, miny              = map_func(-120, 25)
maxx, maxy              = map_func(-50, 50)
heatmap, xedges, yedges = np.histogram2d(y, x, bins=30, range = [[miny, maxy], [minx, maxx]])
extent                  = [xedges[0], xedges[-1], yedges[0], yedges[-1]]

title("Geographic Distribution of Tornado Occurances")

map_func.drawcoastlines(color = "gray")
map_func.drawcountries(color = "gray")
map_func.drawstates(color = "gray")
map_func.drawmapboundary()
map_func.imshow(heatmap, extent=extent, cmap="Blues", interpolation="nearest");
map_func.colorbar();