12 Essential Numpy and Pandas Functions for Efficient Data Analysis

Numpy is a Python extension for scientific computing that provides powerful n‑dimensional arrays, linear algebra, Fourier transforms, random number generation, and seamless integration with C/C++ and Fortran. Beyond these core features, Numpy can serve as a fast, multi‑dimensional container for generic data, enabling rapid interaction with various databases.

Six efficient Numpy functions

1. argpartition() – Finds the indices of the N largest values and optionally sorts them.

x = np.array([12, 10, 12, 0, 6, 8, 9, 1, 16, 4, 6, 0])
index_val = np.argpartition(x, -4)[-4:]
index_val
# array([1, 8, 2, 0], dtype=int64)
np.sort(x[index_val])
# array([10, 12, 12, 16])

2. allclose() – Checks whether two arrays are element‑wise equal within a tolerance, returning a Boolean.

array1 = np.array([0.12,0.17,0.24,0.29])
array2 = np.array([0.13,0.19,0.26,0.31])
np.allclose(array1, array2, 0.1)  # False
np.allclose(array1, array2, 0.2)  # True

3. clip() – Limits the values in an array to a specified interval.

x = np.array([3, 17, 14, 23, 2, 2, 6, 8, 1, 2, 16, 0])
np.clip(x, 2, 5)
# array([3, 5, 5, 5, 2, 2, 5, 5, 2, 2, 5, 2])

4. extract() – Retrieves elements that satisfy a given condition, supporting logical operators.

# Random integers
array = np.random.randint(20, size=12)
cond = np.mod(array, 2) == 1
np.extract(cond, array)
# np.extract(((array < 3) | (array > 15)), array)

5. where() – Returns indices or values where a condition holds, similar to SQL's WHERE clause.

y = np.array([1,5,6,8,1,7,3,6,9])
np.where(y > 5)
# array([2, 3, 5, 7, 8])
np.where(y > 5, "Hit", "Miss")

6. percentile() – Computes the n‑th percentile of array elements along a specified axis.

a = np.array([1,5,6,8,1,7,3,6,9])
print("50th Percentile of a, axis = 0 :", np.percentile(a, 50, axis=0))
# 6.0
b = np.array([[10, 7, 4], [3, 2, 1]])
print("30th Percentile of b, axis = 0 :", np.percentile(b, 30, axis=0))
# [5.1 3.5 1.9]

These six Numpy functions can dramatically simplify common data‑processing tasks. The article then moves on to six equally useful Pandas functions.

Pandas is a Python library that provides fast, flexible, and expressive data structures for handling structured (tabular, multidimensional, heterogeneous) and time‑series data.

Six efficient Pandas functions

1. read_csv(nrows=n) – Reads only the first *n* rows of a CSV file, saving memory and time for large datasets.

import io, requests, pandas as pd
url = "https://raw.github.com/vincentarelbundock/Rdatasets/master/csv/datasets/AirPassengers.csv"
s = requests.get(url).content
# read only first 10 rows
df = pd.read_csv(io.StringIO(s.decode('utf-8')), nrows=10, index_col=0)

2. map() – Maps each value in a Series to another value using a function, dict, or another Series.

dframe = pd.DataFrame(np.random.randn(4, 3), columns=list('ABC'), index=['India','USA','China','Russia'])
changefn = lambda x: "%.2f" % x
dframe['A'].map(changefn)

3. apply() – Applies a user‑defined function along an axis of the DataFrame.

fn = lambda x: x.max() - x.min()
dframe.apply(fn)

4. isin() – Filters rows where column values belong to a specified set.

filter1 = df["value"].isin([112])
filter2 = df["time"].isin([1949.0])
df[filter1 & filter2]

5. copy() – Creates a deep copy of a Pandas object to avoid unintended side‑effects when modifying data.

data = pd.Series(['India','Pakistan','China','Mongolia'])
new = data.copy()
new[1] = 'Changed value'
print(new)
print(data)

6. select_dtypes() – Returns a subset of columns based on their data types.

framex = df.select_dtypes(include="float64")

Finally, the article mentions pivot_table() as a powerful Pandas function for creating pivot tables similar to Excel.

All code examples are available in a Jupyter Notebook on the GitHub repository

https://github.com/kunaldhariwal/12-Amazing-Pandas-NumPy-Functions

.