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

# # NumPy basics

# In[1]:


import numpy as np

# Set seed for reproducibility
np.random.seed(seed=123)


# ## Scalars

# In[2]:


sk_id_curr = np.array(100001)
print ("sk_id_curr: ", sk_id_curr)

# Number of dimensions
print ("sk_id_curr ndim: ", sk_id_curr.ndim)

# Dimensions
print ("sk_id_curr shape:", sk_id_curr.shape)

# Size of elements
print ("sk_id_curr size: ", sk_id_curr.size)

# Data type
print ("sk_id_curr dtype: ", sk_id_curr.dtype)


# ## 1-D Array

# In[3]:


sk_id_curr = np.array([100001, 100005, 100013, 100028, 100038, 100042])
print ("sk_id_curr:", sk_id_curr)
print ("sk_id_curr ndim: ", sk_id_curr.ndim)
print ("sk_id_curr shape:", sk_id_curr.shape)
print ("sk_id_curr size: ", sk_id_curr.size)
print ("sk_id_curr dtype: ", sk_id_curr.dtype)


# ## 2-D array (matrix)

# In[4]:


sk_id_curr = np.array([[100001, 100005], [100013, 100028], [100038, 100042]])
print ("sk_id_curr:", sk_id_curr)
print ("sk_id_curr ndim: ", sk_id_curr.ndim)
print ("sk_id_curr shape:", sk_id_curr.shape)
print ("sk_id_curr size: ", sk_id_curr.size)
print ("sk_id_curr dtype: ", sk_id_curr.dtype)


# ## Functions

# In[5]:


print ("np.zeros((2,2)):", np.zeros((2,2)))
print ("np.ones((2,2)):", np.ones((2,2)))
print ("np.eye((2)):", np.eye((2)))
print ("np.random.random((2,2)):", np.random.random((2,2)))


# # Accessing
# 
# ## Indexing

# In[6]:


sk_id_curr = np.array([100001, 100005, 100013, 100028, 100038, 100042])
print ("sk_id_curr[0]: ", sk_id_curr[0])
sk_id_curr[0] = 0
print ("sk_id_curr: ", sk_id_curr)


# ## Slicing

# In[7]:


sk_id_curr = np.array([[100001, 100005, 111111], [100013, 100028, 222222], [100038, 100042, 333333]])
print (sk_id_curr)
print ("sk_id_curr column 1: ", sk_id_curr[:, 1]) 
print ("sk_id_curr row 0: ", sk_id_curr[0, :]) 
print ("sk_id_curr rows 0,1,2 & cols 1,2:", sk_id_curr[:3, 1:3]) 


# ## Boolean array indexing

# In[8]:


sk_id_curr = np.array([[100001, 100005, 111111], [100013, 100028, 222222], [100038, 100042, 333333]])
print ("sk_id_curr:", sk_id_curr)
print ("sk_id_curr > 111111:", sk_id_curr > 111111)
print ("sk_id_curr[sk_id_curr > 111111]:", sk_id_curr[sk_id_curr > 111111])


# # Array math
# 
# ## Basic math

# In[9]:


x = np.array([[1,2], [3,4]], dtype=np.float64)
y = np.array([[1,2], [3,4]], dtype=np.float64)
print ("x + y:", np.add(x, y)) # or x + y
print ("x - y:", np.subtract(x, y)) # or x - y
print ("x * y:", np.multiply(x, y)) # or x * y


# ## Dot product

# In[10]:


a = np.array([[1,2,3], [4,5,6]], dtype=np.float64) # we can specify dtype
b = np.array([[7,8], [9,10], [11, 12]], dtype=np.float64)
print (a.dot(b))


# ## Sum across a dimension

# In[11]:


x = np.array([[1,2],[3,4]])
print (x)
print ("sum all: ", np.sum(x)) # adds all elements
print ("sum by col: ", np.sum(x, axis=0)) # add numbers in each column
print ("sum by row: ", np.sum(x, axis=1)) # add numbers in each row


# ## Transposing

# In[12]:


print ("x:", x)
print ("x.T:", x.T)


# ## Broadcasting

# In[13]:


x = np.array([[1,2], [3,4]])
y = np.array([5, 6])
z = x + y
print ("z:", z)


# ## Reshaping

# In[14]:


x = np.array([[1,2], [3,4], [5,6]])
print (x)
print ("x.shape: ", x.shape)
y = np.reshape(x, (2, 3))
print ("y.shape: ", y.shape)
print ("y:", y)