#!/usr/bin/env python
# coding: utf-8

# In[1]:


import pandas as pd
data = pd.read_csv('delaware_anomaly.csv')


# In[2]:


data.head()


# In[7]:


data['MERCHANT'].value_counts()


# In[3]:


data.shape


# In[4]:


data.dtypes


# In[13]:


from pycaret.anomaly import *
s = setup(data, session_id = 123,
          ignore_features = ['CAT_DESCR'], high_cardinality_features= ['MERCHANT'],
          numeric_features = ['FISCAL_YEAR', 'FISCAL_PERIOD' ])


# In[14]:


get_config('X').columns


# In[15]:


iforest = create_model('iforest')


# In[16]:


r = assign_model(iforest)
r.head()


# In[21]:


r['Anomaly_Score'].hist(bins=100, figsize=(10,6))


# In[ ]: