A little script to convert a pandas data frame to a JSON object. Is there a better way?

df2json.py

"""
tiny script to convert a pandas data frame into a JSON object
"""

import ujson as json
import pandas
import numpy as np

df = pandas.DataFrame({
    "time" : [1,2,3,4,5],
    "temp" : np.random.rand(5)
})

d = [ 
    dict([
        (colname, row[i]) 
        for i,colname in enumerate(df.columns)
    ])
    for row in df.values
]
return json.dumps(d)

I just found need to do exactly this, except stored as a dict with the row names as keys. For that, I did: dict([(row[0], dict(row[1])) for row in df.iterrows()]). df.iterkv() gets you the transpose of that.

You could also nest a dictionary comprehension.
d = [{colname : row[i] for i, colname in enumerate(df.columns)} for row in df.iterrows()]

This dict comprehension includes only the relevant information:
[{k:df.values[i][v] for v,k in enumerate(df.columns)} for i in range(len(df)) ]

A related pull request:
pandas-dev/pandas#1263

Looks like thats comming in the next version.
http://pandas.pydata.org/pandas-docs/dev/io.html#writing-json

To complete the thread: Convert DataFrame to JSON with pandas.io.

and now since a little while just use:
df.to_json()