Problem Formulation and Solution Overview

In this article, you’ll learn how to iterate over a 1D, 2D and 3D NumPy array using Python.

When working with the NumPy library, you will encounter situations where you will need to iterate through a 1D, 2D and even a 3D array. This article will show you how to accomplish this task.

---

Question: How would we write code to iterate through a 1D, 2D or 3D NumPy array?

We can accomplish this task by one of the following options:

• Method 1: Use a For loop and np.array()
• Method 2: Use a For loop and np.nditer()
• Method 3: Use a For loop and itertools
• Method 4: Use a While loop and np.size
• Method 5: Use a For loop and np.ndenumerate()
• Method 6: Use a For Loop and range()
• Bonus: CSV to np.array()

---

Preparation

Before moving forward, please ensure the NumPy library is installed. Click here if you require instructions.

Then, add the following code to the top of each script. This snippet will allow the code in this article to run error-free.

import numpy as np 

After importing the NumPy library, this library is referenced by calling the shortcode (np).

---

Method 1: Use a For Loop and np.array()

This method uses a For loop combined with np.array() to iterate through a 1D NumPy array. The first five (5) Atomic Numbers from the Periodic Table are generated and displayed for this example.

atomic_els = np.array(np.arange(1,6)) for el in atomic_els: print(el, end=' ')

Above, calls the np.array() function and passes it np.arange(1,6) with two (2) arguments: a start position of one (1) and stop position of five (5) or (stop-1). The result is an array of integers and saves to atomic_els.

[1 2 3 4 5]

Next, a For loop is instantiated and iterates through each element of the 1D NumPy array atomic_els.

The output is sent to the terminal on one (1) line as an additional argument was passed to the print statement (end=' '). This argument replaces the default newline character with a blank space.

1 2 3 4 5

---

Method 2: Uses a For Loop and nditer()

This method uses a for loop combined with np.nditer() to iterate through a NumPy array. The first five (5) Atomic Numbers and Number of Neutrons data from the Periodic Table display for this example.

atomic_data = np.array([np.arange(1,6), [0, 2, 4, 5, 6]]) for dim in atomic_data: for d in dim: print(d, end=' ')

Above, calls the np.array() function and passes np.arange(1,6) as the first argument and an array of associated Number of Neutrons values as the second.

Next, an outer for loop is instantiated. This iterates through each array dimension.

Inside this loop, another for loop is instantiated. The inside loop iterates through each element of the above dimensions.

The output is sent to the terminal on one (1) line as an additional argument is passed to the print statement (end=' '). This argument replaces the default newline character with a blank space.

1 2 3 4 5 0 2 4 5 6

Note: The first dimension is highlighted to easily differentiate the dimensions.

---

Method 3: Use a For Loop and itertools

This method uses a for loop and Python’s built-in itertools library to iterate through a NumPy array. The first three (3) Atomic Numbers, Phase, and Group data from the Periodic Table display for this example.

import itertools atomic_num = np.arange(1,4)

atomic_phase = ['gas', 'gas', 'solid']

atomic_group = [1, 18, 1] for (a, b, c) in itertools.zip_longest(atomic_num, atomic_phase, atomic_group, fillvalue=0): print (a, b, c)

Above, imports the itertools library to use the zip_longest() function.

Then, three (3) arrays are declared containing the relevant data for the first three (3) elements from the Periodic Table. They save respectively to atomic_num, atomic_phase, and atomic_group.

Next, a for loop is instantiated and passed three (3) arguments inside the brackets (a, b, c) which reference the arguments inside the zip_longest(atomic_num, atomic_phase, atomic_group) function.

An additional argument (fillvalue=0) is also passed. This fills in any missing values for lists of unequal lengths with the stated value.

The loop iterates until the end of the arrays are reached, outputting to the terminal for each iteration.

1 gas 1 2 gas 18 3 solid 1

---

Method 4: Use a While Loop and Size

This method uses a while loop combined with np.size to iterate through a NumPy array. The first five (5) Atomic Numbers, Number of Neutrons, and Phase data from the Periodic Table display for this example.

atomic_names = np.array(['Hydrogen', 'Helium', 'Lithium', 'Beryllium', 'Boron'])

el_count = 0 while el_count < atomic_names.size: print(atomic_names[el_count], end=' ') el_count += 1

Above, calls the np.array() function and passes a list containing the first five (5) element names of the Periodic Table. These results save to atomic_names.

Then, a variable, el_count, is declared with a start value of zero (0). This will be a counter variable and lets the While loop know when to stop iterating.

Next, a while loop is instantiated and iterates through the elements of atomic_names. This will continue while the value of el_count remains less than the value of atomic_names.size.

The output is sent to the terminal on one (1) line as an additional argument is passed to the print statement (end=' '). This argument replaces the default newline character with a blank space.

Hydrogen Helium Lithium Beryllium Boron

---

Method 5: Use a For Loop and np.ndenumerate()

This method uses a For loop and np.ndenumerate() to iterate through a NumPy array. The first five (5) Element Names from the Periodic Table display for this example.

atomic_names = np.array(['Hydrogen', 'Helium', 'Lithium', 'Beryllium', 'Boron']) for idx, x in np.ndenumerate(atomic_names): print(idx, x)

Above, calls the np.array() function and passes it an array containing the first five (5) element names from the Periodic Table. The results save to atomic_names.

Next, a for loop is instantiated referencing idx which is the index of the array, and x, which is the array element’s value.

The output is sent to the terminal. Then idx displays a Tuple containing the index value of the array and then the value of x for each iteration.

(0,) Hydrogen (1,) Helium (2,) Lithium (3,) Beryllium (4,) Boron

---

Method 6: Use a For Loop and range()

This method uses a for loop and range()) to iterate through a 3D NumPy array.

nums = np.array([[[1, 2], [3, 4], [5, 6]], [[7, 8], [9, 10], [11, 12]], [[13, 14], [15, 16], [17, 18]], [[19, 20], [21, 22], [23, 23]], [[24, 25], [26, 27], [28, 29]]]) for x in range(0, 5): for y in range(0, 3): for z in range(0, 2): print(nums[x][y][z])

Above declares a 3D NumPy array containing consecutive numbers from 1-29 inclusive. These save to nums.

Next, three (3) for loops are instantiated to loop through and output the contents of nums to the terminal one (1) number per line.

The range() function for each loop is based on the dimensions of the 3D Numpy array, for example:

• The first loop (range(0,5) identifies the total number of tows in the array.

• The next loop (range(0,3) identifies the number of columns in the array.

• The final loop (range(0,2) identifies the elements in each column in the array.

Finally, the output is sent to the terminal (snippet only).

1 2 3 4 5 6 7 8 9 10 ...

---

Bonus: Convert CSV to np.array()

This example reads in a snippet of the Periodic Table as a CSV file. This data is converted to a NumPy array and is output to the terminal.

The CSV file below contains the values of the Atomic Number, Atomic Mass, Number of Neutrons, and the Number of Electrons for the first seven (7) elements from the Periodic Table.

Contents of pt_sample.csv

1,1.007,0 2,4.002,2 3,6.941,4 4,9.012,5 5,10.811,6 6,12.011,6 7,14.007,7

To follow along, save this file as pt_sample.csv and move it into the current working directory.

Before moving forward, please ensure the Pandas library is installed. Click here if you require instructions.

import pandas as pd

import csv file_name = 'pt_sample.csv'

df_data = np.array(list(csv.reader(open(file_name, 'r'), delimiter=','))).astype("float")

print(df_data)

Above, imports the Pandas library and the CSV library. This is needed to work with DataFrames and read in the CSV file.

Then, a filename is declared and saves to file_name.

On the highlighted line, the np.array() function is called and passed the following arguments.

• The csv.reader() function is passed as an argument and this argument passes the open() method to open the specified CSV file in read (r) mode, including the field delimiter (csv.reader(open(file_name, 'r'), delimiter=',')).
• The contents of the CSV file converts to a list.
• This list converts to an np.array().

The results save to df_data and output to the terminal.

[[ 1. 1.007 0. ] [ 2. 4.002 2. ] [ 3. 6.941 4. ] [ 4. 9.012 5. ] [ 5. 10.811 6. ] [ 6. 12.011 6. ] [ 7. 14.007 7. ]]

---

Summary

These seven (7) methods of iterating a NumPy array should give you enough information to select the best one for your coding requirements.

Good Luck & Happy Coding!

---

Programmer Humor – Blockchain

Python Append Row to CSV

To append a row (=dictionary) to an existing CSV, open the file object in append mode using open('my_file.csv', 'a', newline=''). Then create a csv.DictWriter() to append a dict row using DictWriter.writerow(my_dict).

Given the following file 'my_file.csv':

You can append a row (dict) to the CSV file via this code snippet:

import csv # Create the dictionary (=row)

row = {'A':'Y1', 'B':'Y2', 'C':'Y3'} # Open the CSV file in "append" mode

with open('my_file.csv', 'a', newline='') as f: # Create a dictionary writer with the dict keys as column fieldnames writer = csv.DictWriter(f, fieldnames=row.keys()) # Append single row to CSV writer.writerow(row)

After running the code in the same folder as your original 'my_file.csv', you’ll see the following result:

Append Multiple Rows to CSV

Given the following CSV file:

To add multiple rows (i.e., dicts) to an old existing CSV file, iterate over the rows and write each row by calling csv.DictWriter.writerow(row) on the initially created DictWriter object.

Here’s an example (major changes highlighted):

import csv # Create the dictionary (=row)

rows = [{'A':'Z1', 'B':'Z2', 'C':'Z3'}, {'A':'ZZ1', 'B':'ZZ2', 'C':'ZZ3'}, {'A':'ZZZ1', 'B':'ZZZ2', 'C':'ZZZ3'}] # Open the CSV file in "append" mode

with open('my_file.csv', 'a', newline='') as f: # Create a dictionary writer with the dict keys as column fieldnames writer = csv.DictWriter(f, fieldnames=rows[0].keys()) # Append multiple rows to CSV for row in rows: writer.writerow(row)

The resulting CSV file has all three rows added to the first row:

Python Add Row to CSV Pandas

To add a row to an existing CSV using Pandas, you can set the write mode argument to append 'a' in the pandas DataFrame to_csv() method like so: df.to_csv('my_csv.csv', mode='a', header=False).

df.to_csv('my_csv.csv', mode='a', header=False)

For a full example, check out this code snippet:

import pandas as pd # Create the initial CSV data

rows = [{'A':'Z1', 'B':'Z2', 'C':'Z3'}, {'A':'ZZ1', 'B':'ZZ2', 'C':'ZZ3'}, {'A':'ZZZ1', 'B':'ZZZ2', 'C':'ZZZ3'}] # Create a DataFrame and write to CSV

df = pd.DataFrame(rows)

df.to_csv('my_file.csv', header=False, index=False) # Create another row and append row (as df) to existing CSV

row = [{'A':'X1', 'B':'X2', 'C':'X3'}]

df = pd.DataFrame(row)

df.to_csv('my_file.csv', mode='a', header=False, index=False)

The output file looks like this (new row highlighted):

Alternatively, you can open the file in append mode using normal open() function with the append 'a' argument and pass it into the pandas DataFrame to_csv() method.

Here’s an example snippet for copy&paste:

with open('my_csv.csv', 'a') as f: df.to_csv(f, header=False)

Learn More: If you want to learn about 7 Best Ways to Convert Dict to CSV in Python, check out this Finxter tutorial.

Localhost is the web developer’s favorite home. Local development environment is always convenient to develop and debug scripts.

Mailing via a script with in-built PHP function. This may not work almost always in a localhost. You need to have a sendmail program and appropriate configurations.

If the PHP mail() is not working on your localhost, there is an alternate solution to sending email. You can use a SMTP server and send email from localhost and it is the popular choice for sending emails for PHP programmers.

This example shows code to use PHPMailer to send email using SMTP from localhost.

PHP Mail sending script with SMTP

This program uses the PHPMailer library to connect SMTP to send emails. You can test this in your localhost server. You need access to a SMTP server.

Before running this code, configure the SMTP settings to set the following details.

1. Authentication directive and security protocol.
2. Configure SMTP credentials to authenticate and authorize mail sending script.
3. Add From, To and Reply-To addresses using the PHPMailer object.
4. Build email body and add subject before sending the email.

The above steps are mandatory with the PHPMailer mail sending code. Added to that, this mailing library supports many features. Some of them are,

Set the SMTP Debug = 4 in development mode to print the status details of the mail-sending script.

<?php

use PHPMailer\PHPMailer\PHPMailer;

use PHPMailer\PHPMailer\SMTP;

use PHPMailer\PHPMailer\Exception; require_once __DIR__ . '/vendor/phpmailer/src/Exception.php';

require_once __DIR__ . '/vendor/phpmailer/src/PHPMailer.php';

require_once __DIR__ . '/vendor/phpmailer/src/SMTP.php'; $mail = new PHPMailer(true); $mail->SMTPDebug = 0;

$mail->isSMTP();

$mail->Host = 'SET-SMTP-HOST';

$mail->SMTPAuth = true;

$mail->SMTPSecure = "ssl";

$mail->Port = 465; $mail->mailer = "smtp"; $mail->Username = 'SET-SMTP-USERNAME';

$mail->Password = 'SET-SMTP-PASSWORD'; // Sender and recipient address

$mail->SetFrom('SET-SENDER-EMAIL', 'SET-SENDER_NAME');

$mail->addAddress('ADD-RECIPIENT-EMAIL', 'ADD-RECIPIENT-NAME');

$mail->addReplyTo('ADD-REPLY-TO-EMAIL', 'ADD-REPLY-TO-NAME'); // Setting the subject and body

$mail->IsHTML(true);

$mail->Subject = "Send email from localhost using PHP";

$mail->Body = 'Hello World!'; if ($mail->send()) { echo "Email is sent successfully.";

} else { echo "Error in sending an email. Mailer Error: {$mail->ErrorInfo}";

}

?>

Google and Microsoft disabled insecure authentication

Earlier, programmers conveniently used GMail’s SMTP server for sending emails via PHP. Now, less secure APPs configuration in Google is disabled. Google and Microsoft force authentication via OAuth 2 to send email.

There is ill-informed text going around in this topic. People say that we cannot send email via Google SMTP any more. That is not the case. They have changed the authentication method.

IMPORTANT: I have written an article to help you send email using xOAuth2 via PHP. You can use this and continue to send email using Google or other email service providers who force to use OAuth.

Alternate: Enabling PHP’s built-in mail()

1. If you do not have access to an email SMTP server.
2. If you are not able to use xOAuth2 and Google / Microsoft’s mailing service.

In the above situations, you may try to setup your own server in localhost. Yes! that is possible and it will work.

PHP has a built-in mail function to send emails without using any third-party libraries.

The mail() function is capable of simple mail sending requirements in PHP.

In localhost, it will not work without setting the php.ini configuration. Find the following section in your php.ini file and set the sendmail path and related configuration.

Download

↑ Back to Top