In [39]:

    

import pandas as pd
import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt


    

    




ERROR! Session/line number was not unique in database. History logging moved to new session 501

In [40]:

    

%matplotlib inline


    

In [41]:

    

import seaborn as sns


    

In [42]:

    

data_file = '../data/moving_violations.csv'


    

In [43]:

    

df = pd.read_csv(data_file)
df['ticketissuedate'] = pd.to_datetime(df.ticketissuedate)
df['totalpaid'] = df.totalpaid.fillna(0)


    

In [46]:

    

df.info()


    

    




<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3993673 entries, 0 to 3993672
Data columns (total 20 columns):
accidentindicator    object
address_id           float64
agencyid             float64
fineamt              float64
location             object
objectid             int64
penalty1             float64
penalty2             float64
row_                 int64
streetsegid          float64
ticketissuedate      datetime64[ns]
tickettype           object
totalpaid            float64
violationcode        object
violationdesc        object
xcoord               float64
y                    float64
ycoord               float64
filename             object
x                   float64
dtypes: datetime64[ns](1), float64(11), int64(2), object(6)
memory usage: 609.4+ MB

In [ ]:

    




    

In [9]:

    

df.groupby('accidentindicator').totalpaid.mean()


    

    
Out[9]:





accidentindicator
No     94.883731
Yes    50.499074
Name: totalpaid, dtype: float64

In [10]:

    

df.groupby(['accidentindicator', 'tickettype' ]).totalpaid.mean()


    

    
Out[10]:





accidentindicator  tickettype    
No                 Mover Void         0.147887
                   Moving            63.071702
                   No Ticket Type    36.336931
                   Parker Void        0.000000
                   Parking            1.724138
                   Photo             97.735412
Yes                Mover Void         0.000000
                   Moving            50.913755
                   No Ticket Type    27.017391
Name: totalpaid, dtype: float64

In [11]:

    

df['tickettype'] = df.tickettype.fillna('')


    

In [21]:

    

df.columns


    

    
Out[21]:





Index([u'accidentindicator', u'address_id', u'agencyid', u'fineamt',
       u'location', u'objectid', u'penalty1', u'penalty2', u'row_',
       u'streetsegid', u'ticketissuedate', u'tickettype', u'totalpaid',
       u'violationcode', u'violationdesc', u'xcoord', u'y', u'ycoord',
       u'filename', u'x'],
      dtype='object')

In [12]:

    

voided_df = df[df.tickettype.str.contains('Void')]
voided = None


    

In [13]:

    

voided_df = voided_df.groupby(['accidentindicator', 'violationcode']).count().reset_index()


    

In [14]:

    

voided_df['count'] = voided_df.address_id


    

In [15]:

    

t_df = voided_df[['accidentindicator', 'violationcode', 'count']]


    

In [16]:

    

# t_df.sort_values('count', ascending=False)


    

In [17]:

    

sns.set_style("whitegrid")


    

In [ ]:

    




    

In [19]:

    

sns.boxplot(x="tickettype", y="totalpaid", data=df)


    

    
Out[19]:





<matplotlib.axes._subplots.AxesSubplot at 0x12190ab50>






    

In [ ]:

    




    

In [ ]:

    




    

In [20]:

    

sns.boxplot(x="tickettype", y="totalpaid", data=voided_df)


    

    
Out[20]:





<matplotlib.axes._subplots.AxesSubplot at 0x110ab7090>






    

In [29]:

    

ts = pd.Series(df.totalpaid, index=df.ticketissuedate)


    

In [30]:

    

ts = ts.cumsum()


    

In [45]:

    

len(df)


    

    
Out[45]:





3993673

In [ ]:

    




    

In [44]:

    

df.sample(100000).to_csv('moving_violations_sample_large.csv', sep='\t', index=False)


    

In [ ]: