Demonstrate using GeoPandas and geojsonio.py (http://github.com/jwass/geojsonio.py) to view styled data in a few lines. Running the script will open a browser with the final URL: http://geojson.io/#id=gist:/10218442

cambridge_raw_osm.geojson

geopandas-merge.py

import pandas as pd
import geopandas as gpd
import geojsonio

# Set up the simple style spec mapping of highway to color and width
highway = ['motorway', 'trunk', 'primary', 'secondary', 'tertiary', 'residential']
# Colors pasted from colorbrewer2.org, qualitative, 7-class Set1 (removing the yellow color)
colors = """
#e41a1c
#377eb8
#4daf4a
#984ea3
#ff7f00
#a65628
""".split()
widths = range(len(highway), 0, -1)  # Countdown so smallest is width 1

# 'stroke' and 'stroke-width' are the defined properties for line color/width:
# http://github.com/mapbox/simplestyle-spec
style = pd.DataFrame({'stroke': colors,
                      'stroke-width': widths,
                      'highway': highway})
# >>> style
#        highway   stroke  stroke-width
# 0     motorway  #e41a1c             6
# 1        trunk  #377eb8             5
# 2      primary  #4daf4a             4
# 3    secondary  #984ea3             3
# 4     tertiary  #ff7f00             2
# 5  residential  #a65628             1

# Load up the OSM data
df = gpd.read_file('cambridge_raw_osm.geojson')

# Merge the style columns to the data, merge on the highway column. Because
# this is an inner join by default, it will drop any rows where highway is not
# in the style DataFrame.  Because the merge() results in a DataFrame and not
# a GeoDataFrame, we must re-create the GeoDataFrame. Hopefully this can
# be fixed in the future.
df_styled = gpd.GeoDataFrame(df.merge(style, on='highway'))

# Shoot the output to geojson.io and open in a browser.
url = geojsonio.display(df_styled.to_json(na='drop'))
print url

styled_cambridge_osm.geojson

Visit http://geojson.io/#id=gist:/10218442 for the final, styled output.