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

# In[1]:


import pandas as pd
import numpy as np


# In[2]:


# dataframe with one whole column with np.nan and numeric data; 0 included

df1 = pd.DataFrame(data = [[1, np.nan, np.nan, 0],
                          [np.nan, 2, np.nan, 0],
                          [np.nan, np.nan, np.nan, 0],
                          [np.nan, np.nan, np.nan, np.nan]],
                 columns = ["col1", "col2", "col3", "col4"])
df1


# In[3]:


# datatype of each of the series in dataframe
df1.dtypes


# In[4]:


# here the default is bool_only = None, but our dataframe has only floats right now, 
# no column has boolean values, so the result matches the documentation


df1.any(axis = "columns", skipna = False)


# In[5]:


# dataframe with one whole column as np.nan, numeric data; 0 replaced with True or False

df2 = pd.DataFrame(data = [[1, np.nan, np.nan, True],
                          [np.nan, 2, np.nan, True],
                          [np.nan, np.nan, np.nan, False],
                          [np.nan, np.nan, np.nan, np.nan]],
                 columns = ["col1", "col2", "col3", "col4"])
df2


# In[6]:


# Col4 Series as a whole is object 

print(df2.dtypes)
print("\n")

# indivdual elements are still their repective types
print("col4 elements:", [type(each) for each in df2["col4"]])


# In[7]:


# the result we are currently getting when bool_only = None (default) 
df2.any(axis = "columns", skipna = False)


# In[8]:


# the last index 3 which is completely None, now returns True
df2.any(axis = "columns", skipna = False, bool_only = False)


# In[9]:


# we don't have any row that is completely Boolean so my understanding is this should result in empty series,
# but it give result exactly as bool_only = False option

df2.any(axis = "columns", skipna = False, bool_only = True)


# In[10]:


df2.any(axis = "index", skipna = False)


# In[11]:


df2.any(axis = "index", skipna = False, bool_only = False)


# In[12]:


df2.any(axis = "index", skipna = False, bool_only = True)


# In[13]:


# dataframe with np.nan, 0 and boolean

df3 = pd.DataFrame(data = [[np.nan, np.nan, np.nan, True],
                          [np.nan, np.nan, np.nan, False],
                          [0, 0, 0, 0],
                          [np.nan, np.nan, np.nan, np.nan],
                          [False, False, False, False],
                          [True, True, True, True]],
                 columns = ["col1", "col2", "col3", "col4"])
df3


# In[14]:


df3.dtypes


# In[15]:


df3.any(axis = "columns", skipna = False)


# In[16]:


df3.any(axis = "columns", skipna = False, bool_only = False)


# In[17]:


# I expected this to be empty but its giving exact result as above - This I can't really understand
df3.any(axis = "columns", skipna = False, bool_only = True)


# In[18]:


df3.any(axis = "index", skipna = False)


# In[19]:


df3.any(axis = "index", skipna = False, bool_only = False)


# In[20]:


# no Series with all Bool type elements so it returns empty? but then why not in code cell 15? This is baffling!!!!!
df3.any(axis = "index", skipna = False, bool_only = True)