# Install in system python ...
sudo pip install monasca-statsd

# .. or into a virtual env
pip install monasca-statsd

# Import the module.
import monascastatsd as mstatsd

# Create the connection
conn = mstatsd.Connection(host='localhost', port=8125)

# Create the client with optional dimensions
client = mstatsd.Client(connection=conn, dimensions={'env': 'test'})

NOTE: You can also create a client without specifying the connection and it will create the client 
with the default connection information for the monasca-agent statsd processor daemon 
which uses host='localhost' and port=8125.

client = mstatsd.Client(dimensions={'env': 'test'})

# Increment and decrement a counter.
counter = client.get_counter(name='page.views')

counter.increment()
counter += 3

counter.decrement()
counter -= 3

# Record a gauge 50% of the time.
gauge = client.get_gauge('gauge', dimensions={'env': 'test'})

gauge.send('metric', 123.4, sample_rate=0.5)

# Sample a histogram.
histogram = client.get_histogram('histogram', dimensions={'test': 'True'})

histogram.send('metric', 123.4, dimensions={'color': 'red'})

# Time a function call.
timer = client.get_timer()

@timer.timed('page.render')
def render_page():
    # Render things ...
    pass

# Time a block of code.
timer = client.get_timer()

with timer.time('t'):
    # Do stuff
    time.sleep(2)

# Add dimensions to any metric.
histogram = client.get_histogram('my_hist')
histogram.send('query.time', 10, dimensions = {'version': '1.0', 'environment': 'dev'})

http://www.apache.org/licenses/LICENSE-2.0