Compute "less-than count" on df column using pandas

My raw data is as follows:

id_2    column_A
602     1
602     1
602     1
602     1
602     3
602     8
602     15
602
602

My dataframe:

df = pd.read_csv('xxyy.csv')    
df = df.fillna(np.nan)

My requirement is: I have to find the less than count for each value for ex:(for 1 its 0, for 3 its 4,etc)

EDIT:

def func1(value):                   
    return df['column_A'][df['column_A'] < value].count()

for name, df in df.groupby(['id_2']):
    for j in df.index:
       y = func1(df['column_A'].ix[j])   
       print y

what i am getting is:

for 3 its 5, for 8 its 6, for 15 its 4, what should i want is: for 3 it should be 6, for 8 it should be 7, for 15 it should be 8 (included 2 nan)