In [2]:
import numpy as np
In [3]:
import pandas as pd
In [4]:
df = pd.read_csv('ADR_NOV.csv', encoding='latin1', error_bad_lines=False)

C:\Users\teapen\AppData\Local\Continuum\anaconda3\lib\site-packages\IPython\core\interactiveshell.py:2698: DtypeWarning: Columns (21,22,23,24,25,26,27,28,29,31,32,76,77,78,79,80,81,82,83,84,85,86,87,88,89) have mixed types. Specify dtype option on import or set low_memory=False.
  interactivity=interactivity, compiler=compiler, result=result)
In [5]:
df_cleaned = df[pd.to_numeric(df['INTERACTION_ID'], errors='coerce').notnull()]
In [6]:
desired_rt = [np.nan, 'HANDLED', 'ABANDONED']
In [7]:
df_cleaned = df_cleaned[df_cleaned.RECORD_TYPE.isin(desired_rt)]
In [11]:
df_cleaned['INTERACTION_ID_COUNT'] = 1
In [15]:
grouped = df_cleaned.INTERACTION_ID_COUNT.groupby(df.INTERACTION_ID)
In [23]:
iic_series = grouped.sum()
In [36]:
iic_series = iic_series[1:]
In [39]:
iic_series.mean()
Out[39]:
1.8310411508503064