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

# In[1]:


get_ipython().run_line_magic('matplotlib', 'inline')
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from datetime import datetime
import seaborn as sb


# In[76]:


hospitalized = pd.read_csv('data/hospitalized.csv', index_col=0)


# In[3]:


hospitalized.shape


# ## Virus-negative individuals

# In[39]:


hospitalized['pcr_growth'] = (hospitalized[hospitalized.columns[hospitalized.columns.str.startswith('pcr_result')]]
                                     .sum(1)==0).replace({False:'Growth', True:'No Growth'})


# In[45]:


by_growth = hospitalized.groupby('pcr_growth')


# Plot of age by PCR growth status

# In[42]:


hospitalized.hist('age_months', by='pcr_growth', sharex=True)


# Plot of length of stay by PCR growth

# In[43]:


hospitalized.hist('length_of_stay', by='pcr_growth', sharex=True)


# Higher sepsis and blood culture rates for no growth individuals

# In[46]:


by_growth[['adm_pneumo', 'adm_bronchopneumo', 'adm_sepsis', 'adm_bronchiolitis', 'blood_culture']].mean()


# Bacteria positives

# In[49]:


by_growth[['blood_acintobacter',
 'blood_alcaligenese',
 'blood_candida',
 'blood_ecoli',
 'blood_klebsiella',
 'blood_pneumo',
 'blood_mening',
 'blood_staph',
 'blood_strep',
 'blood_other_gram_neg']].mean().round(3)


# In[68]:


by_sepsis = hospitalized.groupby('adm_sepsis')


# In[74]:


pcr_lookup = {'pcr_result___1': 'RSV',
'pcr_result___2': 'HMPV',
'pcr_result___3': 'flu A',
'pcr_result___4': 'flu B',
'pcr_result___5': 'rhino',
'pcr_result___6': 'PIV1',
'pcr_result___7': 'PIV2',
'pcr_result___8': 'PIV3',
'pcr_result___13': 'H1N1',
'pcr_result___14': 'H3N2',
'pcr_result___15': 'Swine',
'pcr_result___16': 'Swine H1',
'pcr_result___17': 'flu C',
'pcr_result___18': 'Adeno'}

hospitalized['RSV'] = hospitalized.pcr_result___1.astype(int)
hospitalized['HMPV'] = hospitalized.pcr_result___2.astype(int)
hospitalized['Rhino'] = hospitalized.pcr_result___5.astype(int)
hospitalized['Influenza'] = (hospitalized.pcr_result___3 | hospitalized.pcr_result___4).astype(int)
hospitalized['Adeno'] = hospitalized.pcr_result___18.astype(int)
hospitalized['PIV'] = (hospitalized.pcr_result___6 | hospitalized.pcr_result___7 | hospitalized.pcr_result___8).astype(int)
hospitalized['No virus'] = (hospitalized[list(pcr_lookup.keys())].sum(1) == 0).astype(int)


# Risk factors by sepsis status

# In[75]:


by_sepsis[['age_months', 'icu', 'length_of_stay', 'oxygen', 'no_growth', 'death',
          'RSV', 'Adeno', 'PIV', 'Influenza', 'Rhino']].describe()