Try the "column" keyword. For example:

    prices = history(4, '1d', 'close_price')  
    changes = prices.pct_change().dropna()  
    sorted_changes = changes.sort(columns=changes.columns[0])

Instead of selecting the sort column with a positional index, you can also use a column name. Details here: http://pandas.pydata.org/pandas-docs/dev/generated/pandas.DataFrame.sort.html

Thanks, but that did not seem to sort as intended either, see output below:

2015-01-05PRINTSecurity(2 [AA])         -0.057311  
Security(24 [AAPL])      -0.027905  
Security(26892 [HLF])    -0.121938  
Security(39840 [TSLA])   -0.042041  
Name: 2015-01-05 00:00:00+00:00, dtype: float64  
2015-01-06PRINTSecurity(2 [AA])          0.006019  
Security(24 [AAPL])      -0.000188  
Security(26892 [HLF])    -0.078532  
Security(39840 [TSLA])    0.005236  
Name: 2015-01-06 00:00:00+00:00, dtype: float64  

I'm sorry, I misunderstood what you wanted to do. Here's a fix:

def initialize(context):  
    context.secs = [sid(2), sid(24), sid(26892), sid(39840)]

def handle_data(context, data):  
    prices = history(2, '1d', 'close_price')  
    changes = prices.pct_change().dropna().iloc[0]  
    sorted_changes = changes.order()  
    log.info('\n{sc}'.format(sc=sorted_changes))  

Output:

2015-01-05handle_data:8INFO  
Security(26892 [HLF])    -0.121938  
Security(2 [AA])         -0.057311  
Security(39840 [TSLA])   -0.042041  
Security(24 [AAPL])      -0.027905  
Name: 2015-01-05 00:00:00+00:00, dtype: float64  
2015-01-06handle_data:8INFO  
Security(26892 [HLF])    -0.078532  
Security(24 [AAPL])      -0.000188  
Security(39840 [TSLA])    0.005236  
Security(2 [AA])          0.006019  
Name: 2015-01-06 00:00:00+00:00, dtype: float64  

Thank you! I will need to figure out why order() works versus sort() in this instance.

Appreciate the help.