Loop Data Structures
Iterating Over Dictionaries
To iterate through a dictionary's key-value pairs, use the items() method. This ensures each iteration retrieves a key and its corresponding value. Note that dictionaries in Python (prior to version 3.7) do not guarantee a fixed order of iteration.
world = {
"Afghanistan": 38928346,
"Belgium": 11589623,
"China": 1439323776,
"Denmark": 5822763,
"Ethiopia": 114963588,
"France": 65273511
}
for k, v in world.items():
print(f"Key: {k}, Value: {v}")
Where:
- The variable
kandvare arbitrary - Order of iteration may vary in Python versions before 3.7.
Output:
Key: Afghanistan, Value: 38928346
Key: Belgium, Value: 11589623
Key: China, Value: 1439323776
Key: Denmark, Value: 5822763
Key: Ethiopia, Value: 114963588
Key: France, Value: 65273511
Iterating Over NumPy Arrays
NumPy arrays can be iterated over using a basic for loop. For 2D arrays, a simple for loop yields entire sub-arrays, not individual elements.
import numpy as np
np_height = np.array([1.75, 1.80, 1.65, 1.70, 1.68])
np_weight = np.array([68, 74, 59, 72, 65])
meas = np.array([np_height, np_weight])
for row in meas:
print(row)
Output:
[1.75 1.8 1.65 1.7 1.68]
[68. 74. 59. 72. 65. ]
To access each element of a 2D array, use the nditer() function, which efficiently iterates over every value.
# Iterating with nditer()
for element in np.nditer(meas):
print(element)
Output:
1.75
1.8
1.65
1.7
1.68
68.0
74.0
59.0
72.0
65.0
Iterating Through Pandas DataFrame
Consider the sample Pandas dataframe below:
import pandas as pd
data = {
"country": ["Brazil", "Russia", "India", "China", "South Africa"],
"capital": ["Brasília", "Moscow", "New Delhi", "Beijing", "Pretoria"],
"area": [8.5, 17.1, 3.3, 9.6, 1.2],
"population": [211, 144, 1380, 1393, 58]
}
brics = pd.DataFrame(data)
brics.index = ["BR", "RU", "IN", "CH", "SA"]
print(brics)
Output:
country capital area population
BR Brazil Brasília 8.5 211
RU Russia Moscow 17.1 144
IN India New Delhi 3.3 1380
CH China Beijing 9.6 1393
SA South Africa Pretoria 1.2 58
To iterate over each row, use iterrows() which provides the row label and data as a Pandas Series.
for label, row in brics.iterrows():
print(label)
print(row)
Output:
BR
country Brazil
capital Brasília
area 8.5
population 211
Name: BR, dtype: object
RU
country Russia
capital Moscow
area 17.1
population 144
Name: RU, dtype: object
IN
country India
capital New Delhi
area 3.3
population 1380
Name: IN, dtype: object
CH
country China
capital Beijing
area 9.6
population 1393
Name: CH, dtype: object
SA
country South Africa
capital Pretoria
area 1.2
population 58
Name: SA, dtype: object
Selective Print
We can also use subsetting to print selected columns, let's say the labels and the 'capital' column:
for label, row in brics.iterrows():
print(label + ": " + row["capital"])
Output:
BR: Brasília
RU: Moscow
IN: New Delhi
CH: Beijing
SA: Pretoria
Adding a New Column
You can calculate a new column, such as the length of each country name, and add it to the DataFrame.
for label, row in brics.iterrows():
brics.loc[label, "name_length"] = len(row["country"])
print(brics)
Output:
country capital area population name_length
BR Brazil Brasília 8.5 211 6.0
RU Russia Moscow 17.1 144 6.0
IN India New Delhi 3.3 1380 5.0
CH China Beijing 9.6 1393 5.0
SA South Africa Pretoria 1.2 58 12.0
This is okay for small datasets but will be extremely problematic for larger datasets. or better performance, use apply() to calculate the name_length column without needing a for loop.
brics["name_length"] = brics["country"].apply(len)
print(brics)
This will yield the same output:
country capital area population name_length
BR Brazil Brasília 8.5 211 6
RU Russia Moscow 17.1 144 6
IN India New Delhi 3.3 1380 5
CH China Beijing 9.6 1393 5
SA South Africa Pretoria 1.2 58 12