def gray_plot(data, min=0, max=1, name=""):
reshape = importr('reshape')
gg = ggplot2.ggplot(reshape.melt(data, id_var=['x', 'y']))
pg = gg + ggplot2.aes_string(x='L1',y='L2')+ \
ggplot2.geom_tile(ggplot2.aes_string(fill='value'))+ \
ggplot2.scale_fill_gradient(low="black", high="white",limits=FloatVector((min,max)))+ \
ggplot2.coord_equal() + ggplot2.scale_x_continuous(name)
return pg
def gray_plot(data, min=0, max=1, name=""):
reshape = importr('reshape')
gg = ggplot2.ggplot(reshape.melt(data,id_var=['x','y']))
pg = gg + ggplot2.aes_string(x='L1',y='L2')+ \
ggplot2.geom_tile(ggplot2.aes_string(fill='value'))+ \
ggplot2.scale_fill_gradient(low="black", high="white",limits=FloatVector((min,max)))+ \
ggplot2.coord_equal() + ggplot2.scale_x_continuous(name)
return pg
## note that different from R dot . is not valid for Python variable names!
IL_railroads = robjects.r('IL.railroads')
IL_final = robjects.r('IL.final')
## import device driver from R with importr to plot to PNG
## we can then call any function in the grdevices package
grdevices = importr('grDevices')
grdevices.png(file='/Users/user/Downloads/data/mapplot.png', width=1300, height=1000)
## plot the map
## note that the order matters when we add another layer in ggplot (here IL_railroads): first aes, then data, that's different from R
## (see http://permalink.gmane.org/gmane.comp.python.rpy/2349)
## note that we use dictionary to set the opts to be able to set options as keywords, for example legend.key.size
p_map = ggplot2.ggplot(IL_final) + \
ggplot2.geom_polygon(ggplot2.aes(x = 'long', y = 'lat', group = 'group', color = 'ObamaShare', fill = 'ObamaShare')) + \
ggplot2.scale_fill_gradient(high = 'blue', low = 'red') + \
ggplot2.scale_fill_continuous(name = "Obama Vote Share") + \
ggplot2.scale_colour_continuous(name = "Obama Vote Share") + \
ggplot2.opts(**{'legend.position': 'left', 'legend.key.size': robjects.r.unit(2, 'lines'), 'legend.title' : ggplot2.theme_text(size = 14, hjust=0), \
'legend.text': ggplot2.theme_text(size = 12), 'title' : "Obama Vote Share and Distance to Railroads in IL", \
'plot.title': ggplot2.theme_text(size = 24), 'plot.margin': robjects.r.unit(robjects.r.rep(0,4),'lines'), \
'panel.background': ggplot2.theme_blank(), 'panel.grid.minor': ggplot2.theme_blank(), 'panel.grid.major': ggplot2.theme_blank(), \
'axis.ticks': ggplot2.theme_blank(), 'axis.title.x': ggplot2.theme_blank(), 'axis.title.y': ggplot2.theme_blank(), \
'axis.title.x': ggplot2.theme_blank(), 'axis.title.x': ggplot2.theme_blank(), 'axis.text.x': ggplot2.theme_blank(), \
'axis.text.y': ggplot2.theme_blank()} ) + \
ggplot2.geom_line(ggplot2.aes(x='long', y='lat', group='group'), data=IL_railroads, color='grey', size=0.2) + \
ggplot2.coord_equal()
p_map.plot()
## add the scatterplot
IL_final = robjects.r('IL.final')
## import device driver from R with importr to plot to PNG
## we can then call any function in the grdevices package
grdevices = importr('grDevices')
grdevices.png(file='/Users/user/Downloads/data/mapplot.png',
width=1300,
height=1000)
## plot the map
## note that the order matters when we add another layer in ggplot (here IL_railroads): first aes, then data, that's different from R
## (see http://permalink.gmane.org/gmane.comp.python.rpy/2349)
## note that we use dictionary to set the opts to be able to set options as keywords, for example legend.key.size
p_map = ggplot2.ggplot(IL_final) + \
ggplot2.geom_polygon(ggplot2.aes(x = 'long', y = 'lat', group = 'group', color = 'ObamaShare', fill = 'ObamaShare')) + \
ggplot2.scale_fill_gradient(high = 'blue', low = 'red') + \
ggplot2.scale_fill_continuous(name = "Obama Vote Share") + \
ggplot2.scale_colour_continuous(name = "Obama Vote Share") + \
ggplot2.opts(**{'legend.position': 'left', 'legend.key.size': robjects.r.unit(2, 'lines'), 'legend.title' : ggplot2.theme_text(size = 14, hjust=0), \
'legend.text': ggplot2.theme_text(size = 12), 'title' : "Obama Vote Share and Distance to Railroads in IL", \
'plot.title': ggplot2.theme_text(size = 24), 'plot.margin': robjects.r.unit(robjects.r.rep(0,4),'lines'), \
'panel.background': ggplot2.theme_blank(), 'panel.grid.minor': ggplot2.theme_blank(), 'panel.grid.major': ggplot2.theme_blank(), \
'axis.ticks': ggplot2.theme_blank(), 'axis.title.x': ggplot2.theme_blank(), 'axis.title.y': ggplot2.theme_blank(), \
'axis.title.x': ggplot2.theme_blank(), 'axis.title.x': ggplot2.theme_blank(), 'axis.text.x': ggplot2.theme_blank(), \
'axis.text.y': ggplot2.theme_blank()} ) + \
ggplot2.geom_line(ggplot2.aes(x='long', y='lat', group='group'), data=IL_railroads, color='grey', size=0.2) + \
ggplot2.coord_equal()
p_map.plot()
## add the scatterplot
temp_change = temp_change.rename(columns={'passenger_count':'percent_change'})
pickups_change = pd.concat([pickups_change, temp_change['percent_change']], axis=1)
pickups_change = pickups_change.dropna()
# Find top 10 neighborhoods with largest percent change
print(tabulate(pickups_change[['percent_change', 'nbhd',
'borough']].drop_duplicates().sort(['percent_change'], ascending=False).head(10),
tablefmt='pipe', headers='keys', showindex=False))
pickups_change['percent_change'] = np.log1p(pickups_change['percent_change'] + abs(min(pickups_change['percent_change'])))
pickups_change = pd.merge(pickups_change, nbhd_borders, how='right', on=['nbhd']).dropna()
p5 = ggplot2.ggplot(pandas2ri.py2ri(pickups_change)) + \
ggplot2.aes_string(x='lon', y='lat', group='nbhd', fill='percent_change') + \
ggplot2.geom_polygon() + \
ggplot2.scale_fill_gradient(low='blue', high='green') + \
ggplot2.theme(legend_position='bottom') + \
ggplot2.labs(x='', y='', title='Change In Annual # of Pickups (2010 - 2015)', fill='Percent Change\n(Log-Scale)')
p5.save('./plots/2010_2015_percent_change.png', width=5, height=6)
#################################################################
#################################################################
################ MODEL FITTING ##################################
#################################################################
#################################################################
# Do some quick benchmarks (predict 2015 from 2014 data)
from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import RandomizedSearchCV
from scipy.stats import randint as sp_randint
