TAKEAWAYS
Accounting is no longer just about balancing books and preparing reports. The industry is undergoing a transformation, moving away from manual tasks to data-driven insights powered by machine learning (ML). At the heart of this change is Python, a programming language that bridges automation, analytics, and ML, making it highly accessible and practical for accountants.
By understanding the philosophy of programming, accountants can gain a valuable perspective on how programming bridges human logic with machine capabilities. Accountants, like programmers, work with structured systems – financial data, regulatory frameworks, and organisational processes. Learning Python helps accountants enhance their problem-solving capabilities, automate repetitive tasks, and optimise workflows for greater efficiency and impact.
The process of transitioning from pseudocode or pseudo-English to working code in software development follows a structured approach, as depicted below:
This process begins with understanding the problem by identifying inputs, outputs, and business logic. Next, pseudocode is written to outline a high-level algorithm, focusing on the logic and sequence of operations. A suitable programming language, such as Python, is then selected based on the project’s requirements and constraints.
The pseudocode is systematically translated into syntactically correct code, using appropriate data structures and syntax. When transitioning from pseudocode to actual code, the clarity and structure of pseudocode make the coding process straightforward. For example, consider the task of calculating the sum of numbers in a list:
Pseudocode:
DEFINE a function to calculate the sum of numbers
INITIALIZE a variable total to 0
FOR each number in the list
ADD the number to total
RETURN total
Python code:
def calculate_sum(numbers):
total = 0 # Initialize total to 0
for num in numbers: # Loop through each number in the list
total += num # Add the current number to total
return total # Return the calculated sum
Here, the transformation from pseudocode to Python is seamless because the pseudocode clearly outlines the steps to solve the problem – defining a function, initialising variables, looping through the list, and performing a cumulative operation. By breaking down the problem logically in pseudocode, the process of writing syntactically correct Python code becomes intuitive and systematic. This highlights the importance of understanding the problem and planning its solution before diving into the coding stage.
The followup stage of incremental testing ensures each segment works as intended, and components are integrated and optimised for clarity and efficiency. The final application is validated through comprehensive testing, ensuring it meets all requirements. Lastly, documentation and maintenance plans ensure the system remains understandable and future-ready.
For accountants, Python offers a practical way to automate tasks, analyse financial data, and integrate ML models to drive smarter decision-making. The following sections illustrate how Python is transforming the accounting profession using pseudocode to express some of the tasks.
1. Automating repetitive tasks
Traditionally, accountants spend significant time on repetitive tasks like data entry, reconciliations, and report generation. Python simplifies these processes through automation, freeing up time for more strategic work.
Example: Accountants can automate Excel-based tasks, reducing errors and saving hours of effort.
Pseudocode:
LOAD the Excel workbook
SELECT the worksheet
FOR each row in the worksheet (starting from row 2):
READ the current value in column B
CALCULATE the new value as current value * 1.1
WRITE the new value into column C
SAVE the workbook
PRINT "Excel updated successfully"
2. Enhancing data analytics
Python’s ability to handle large datasets efficiently gives accountants a competitive edge. Tools like Pandas and NumPy allow professionals to clean, analyse, and interpret financial data, going far beyond the limitations of spreadsheets.
Example: Identifying trends in revenue or spotting anomalies in expense data becomes faster and more accurate.
Pseudocode:
LOAD the financial data from the CSV file
GROUP data by the "Year" column
SUM the "Revenue" column for each group
PRINT "Total Revenue by Year:"
PRINT the grouped totals
OPTIONAL: CREATE a bar chart of revenue grouped by year
3. Introducing ML for predictive insights
ML is no longer exclusive to tech giants; it’s increasingly relevant for accounting, enabling accountants to build predictive models tailored to their needs.
Example: Forecasting cash flow or predicting client payment behaviour becomes more accurate with ML algorithms.
Pseudocode:
LOAD the payment data from the CSV file
SELECT the features: "InvoiceAmount", "DaysLate", "ClientType"
SET the target variable as "PaymentMade" (1 for paid, 0 for not paid)
SPLIT the data into training and testing sets (80% training, 20% testing)
INITIALIZE an ML algorithm
TRAIN the algorithm on the training set
PREDICT the outcomes on the testing set
CALCULATE and PRINT the accuracy of the predictions
4. Transforming audit and compliance
Python-powered ML tools help auditors detect fraud and ensure compliance by identifying patterns and anomalies in vast datasets – tasks that would be impossible manually.
Example: An ML model could flag unusual transactions for further review, enhancing audit quality and efficiency.
Pseudocode:
LOAD the transaction data from the CSV file
INITIALIZE an ML algorithm
TRAIN the model on the "TransactionAmount" column
ADD a new column "Outlier" to the dataset:
- SET value to -1 for outliers
- SET value to 1 for normal transactions
FILTER rows where "Outlier" is -1 (anomalies)
PRINT the filtered rows (unusual transactions)
With Python and ML handling routine and complex analytical tasks, accountants can focus on strategic decision-making and advisory roles. This shift enables accountants to evolve from being task-driven to insight-driven, providing greater value to their organisations. Python is not just changing how accountants work, it’s redefining what accounting can achieve. By combining automation with ML, Python empowers accountants to work smarter, reduce errors, and deliver actionable insights. For professionals ready to adapt, Python represents an opportunity to become indispensable in a rapidly evolving industry.
Dr Lim Thou Tin is Principal Consultant, Decision Modelling Systems.
© 2024 Institute of Singapore Chartered Accountants
