NumPy: dtype Data Types

np.int32    # 32-bit signed integer
np.float64  # 64-bit floating point
np.bool_    # boolean
np.str_     # unicode string
np.object_  # Python object

import numpy as np

a = np.array([1, 2, 3])
print(a.dtype)       # int64 (on 64-bit platforms)
print(a.dtype.kind)  # 'i'
print(a.dtype.itemsize)  # 8

> **Output:** Run NumPy 2.x in your local Python environment to see the ndarray output. The Piston server does not have NumPy pre-installed — install it locally (`pip install numpy`) and compare. Actual values may vary by NumPy version and random seed.

import numpy as np

# By NumPy type object
a = np.array([1, 2, 3], dtype=np.float32)

# By string
b = np.array([1, 2, 3], dtype='float32')

# By single-character code
c = np.array([1, 2, 3], dtype='f4')  # float32: 'f' + 4 bytes

import numpy as np

# astype returns a new array with the target dtype
a = np.array([1, 2, 3], dtype=np.int32)
b = a.astype(np.float64)  # [1. 2. 3.]

# Overflow: values beyond the target range wrap around
c = np.array([100, 200, 300], dtype=np.uint8)
# [100, 200, 44]  — 300 - 256 = 44

# Float to int truncates
d = np.array([1.9, 2.5, 3.1])
print(d.astype(np.int32))  # [1 2 3]

import numpy as np

# Define a structured dtype
dt = np.dtype([('name', 'U10'), ('age', 'i4'), ('salary', 'f8')])

# Create array with structured dtype
employees = np.array([
    ('Alice', 30, 75000.0),
    ('Bob', 25, 62000.0),
    ('Charlie', 35, 85000.0)
], dtype=dt)

print(employees['name'])   # ['Alice' 'Bob' 'Charlie']
print(employees['salary'].mean())  # 74000.0

> **Output:** Run NumPy 2.x in your local Python environment to see the ndarray output. The Piston server does not have NumPy pre-installed — install it locally (`pip install numpy`) and compare. Actual values may vary by NumPy version and random seed.

import numpy as np

# int8 range: -128 to 127
vals = np.array([100, 120, 127, 128, 130, 200], dtype=np.int8)
print("int8 values:", vals)
# [100  120  127 -128 -126  -56]

# uint8 range: 0 to 255
vals_u = np.array([200, 255, 256, 300], dtype=np.uint8)
print("uint8 values:", vals_u)
# [200 255   0  44]

> **Output:** Run NumPy 2.x in your local Python environment to see the ndarray output. The Piston server does not have NumPy pre-installed — install it locally (`pip install numpy`) and compare. Actual values may vary by NumPy version and random seed.

> **Output:** Run NumPy 2.x in your local Python environment to see the ndarray output. The Piston server does not have NumPy pre-installed — install it locally (`pip install numpy`) and compare. Actual values may vary by NumPy version and random seed.

import numpy as np

dt = np.dtype([('name', 'U10'), ('dept', 'U15'), ('salary', 'f8'), ('years', 'i2')])

data = np.array([
    ('Alice', 'Engineering', 95000, 5),
    ('Bob', 'Marketing', 72000, 3),
    ('Charlie', 'Engineering', 110000, 8),
    ('Diana', 'Sales', 68000, 2),
], dtype=dt)

# Query by department
eng = data[data['dept'] == 'Engineering']
print("Engineering avg salary:", eng['salary'].mean())

# Sort by salary
sorted_data = np.sort(data, order='salary')
print("Top earner:", sorted_data[-1]['name'], sorted_data[-1]['salary'])

> **Output:** Run NumPy 2.x in your local Python environment to see the ndarray output. The Piston server does not have NumPy pre-installed — install it locally (`pip install numpy`) and compare. Actual values may vary by NumPy version and random seed.

import numpy as np

# Compare memory usage of different dtypes
arr_int64 = np.ones(1000000, dtype=np.int64)
arr_float32 = np.ones(1000000, dtype=np.float32)
arr_int8 = np.ones(1000000, dtype=np.int8)

print(f"int64:  {arr_int64.nbytes / 1e6:.1f} MB")
print(f"float32: {arr_float32.nbytes / 1e6:.1f} MB")
print(f"int8:   {arr_int8.nbytes / 1e6:.1f} MB")

# Convert between types
converted = arr_float32.astype(np.int32)
print(f"Converted dtype: {converted.dtype}")