NumPy with Pandas — How They Work Together

import numpy as np
import pandas as pd

df = pd.DataFrame({'a': [1, 2, 3], 'b': [4.0, 5.0, 6.0]})

# .to_numpy() is preferred over .values
arr = df.to_numpy()
print(arr)
print(type(arr))  # numpy.ndarray
print(arr.dtype)  # float64 — upcast to accommodate both int and float

# Single column to array
col = df['a'].to_numpy()
print(col)  # [1 2 3]

# NumPy array back to DataFrame
back = pd.DataFrame(arr, columns=['a', 'b'])
print(back)

import numpy as np
import pandas as pd

s = pd.Series([1.0, 4.0, 9.0, 16.0])

# NumPy ufuncs work directly on Series — preserve the index
print(np.sqrt(s))
# 0    1.0
# 1    2.0
# 2    3.0
# 3    4.0

df = pd.DataFrame({'x': [1, 2, 3], 'y': [10, 20, 30]})
print(np.log(df))  # works on entire DataFrame

import numpy as np
import pandas as pd

df = pd.DataFrame(np.random.randn(10000, 50))

# Pandas matrix multiply — slower due to overhead
result_pd = df.values @ df.values.T

# Pure NumPy — faster for large arrays
arr = df.to_numpy()
result_np = arr @ arr.T

print(result_np.shape)  # (10000, 10000)
print(np.allclose(result_pd, result_np))  # True