# In[1]
name=['Alice','Bob','Cathy','Doug']
age=[25,45,37,19]
weight=[55.0,85.5,68.0,61.5]

---

# In[2]
data=np.zeros(4,dtype={'names':('name','age','weight'),'formats':('U10,'i4','f8')})
print(data.dtype)

# Out[2]
[('name', '<U10'), ('age', '<i4'), ('weight', '<f8')]

• U10 means Unicode string of maximum length 10, i4 means 4 byte integer, and f8 means 8 byte float.
• We can fill the array with our lists of values.

# In[3]
data['name']=name
data['age']=age
data['weight']=weight
print(data)

# Out[3]
[('Alice', 25, 55. ) ('Bob', 45, 85.5) ('Cathy', 37, 68. )
 ('Doug', 19, 61.5)]

---

# In[4]
data['name']

# Out[4]
array(['Alice' 'Bob' 'Cathy' 'Doug'], dtype='<U10')

• We can also using boolean masking

# In[5]
data[data['age']<30]['name']

# Out[5]
array(['Alice', 'Doug'], dtype='<U10')

Exploring Structured Arrays Creation

# In[6]
np.dtype({'names':('name','age','weight'),'formats':('U10','i4','f8')})

# Out[6]
dtype([('name', '<U10'), ('age', '<i4'), ('weight', '<f8')])

• numerical types can be specified using Python types or Numpy dtypes instead like this

# In[7]
np.dtype({'names':('name','age','weight'),'formats':((np.str_,10),int,np.float32)})

# Out[7]
dtype([('name', '<U10'), ('age', '<i8'), ('weight', '<f4')])

• A compound type can also be specified as a list of tuples.

# In[8]
np.dtype([('name','S10'),('age','i4'),('weight','f8')])

# Out[8]
dtype([('name', 'S10'), ('age', '<i4'), ('weight', '<f8')])

• If the names of the types do not matter to you, you can specify the types alone in a comma-separated string

# In[9]
np.dtype('S10,i4,f8')

# Out[9]
dtype([('f0', 'S10'), ('f1', '<i4'), ('f2', '<f8')])

Numpy data types

Character	Description	Example
b	Byte	np.dtype('b')
i	Signed integer	np.dtype('i4') == np.int32
u	Unsigned integer	np.dtype('u1') == np.uint8
f	Floating point	np.dtype('f8') == np.int64
c	Complex floating point	np.dtype('c16') == np.complex128
S,a	String	np.dtype('S5')
U	Unicode string	np.dtype('U') == np.str_
V	Raw data(void)	np.dtype('V') == np.void

More Advanced Compound Types

• We can create a type where each element contains an array or matrix of values.

# In[10]
tp=np.dtype([('id','i8'),('mat','f8',(3,3))])
X=np.zeros(1,dtype=tp)
print(X[0])
print(X['mat'][0])

# Out[10]
(0, [[0., 0., 0.], [0., 0., 0.], [0., 0., 0.]])
[[0. 0. 0.]
 [0. 0. 0.]
 [0. 0. 0.]]

• Each element in the X array consists of an id and a 3×3 matrix.

Record Arrays: Structured Arrays with a Twist

• Numpy also provides record arrays (like np.recarray), which are almost identical to the structured arrays just described
• However, field can be accessed as attributes rather than as dictionary keys.

# In[11]
data['age']

# Out[11]
array([25, 45, 37, 19], dtype=int32)

• If we view our data as a record array instead, we can access like this

# In[12]
data_rec=data.view(np.recarray)
data_rec.age

# Out[12]
array([25, 45, 37, 19], dtype=int32)

• The In[12] is that for record arrays, there is some extra overhead involved in accessing the fields.