CalcTree’s Calculation Verification Process

This article outlines CalcTree’s calculation verification process that ensures our engineering calculation templates are accurate, efficient, and reliable.
﻿
The verification process is done in three stages, technical review, page content review and performance review. Each stage ensures a key component of the original calculator is reviewed thoroughly to make sure everything is working correctly and up to industry standards. 
﻿
﻿
﻿
Where required, steps are returned to and iterated on to ensure correctness and robustness. This is particularly relevant in the final performance review. 
﻿
1 - Technical ReviewIn stage 1, the technical review focuses on the core aspects of the calculations themselves. It involves verifying accuracy and ensuring everything is working as it should.
1.1 Review key documentation The first step in the process is to identify and understand all relevant industry information related to the calculation. In most cases this will be code documentation and/or manufacturer's specification documents, for calculations related to specific products. While this is done in the initial build process of calculations, we re-check again in the verification process to ensure we haven’t missed anything.
1.1.1 Identify Relevant Codes: CalcTree identifies applicable engineering codes and standards for each calculation to ensure compliance with industry guidelines and safety requirements. 
1.1.2 Collect Manufacturers' Documentation: If applicable, gather manufacturers' datasheets and technical specifications to understand the products and components involved.
1.1.3 Collect auxiliary resources: The team reviews appendices, online materials and other resources (e.g. journal articles) which may impact a calculator
1.1.4 Review all documentation and build notes: Our team carefully reviews the collated documentation from 1.1.1 to 1.1.3  to gain a full understanding of the guidelines and specifications. We note down key information where applicable for usage in later stages.
﻿
1.2 Calculation source reviewIn most cases, our calculators will use an integrated source such as Python or a spreadsheet to perform the majority of calculations. This step in the review focuses on the accuracy, organisation, and presentation of these sources to ensure that they are not only correct but also user-friendly, accessible and not a black box.
1.2.1 Verify Correctness: CalcTree ensures the correctness of calculations by cross-checking formulas, algorithms, and data inputs with documentation and findings from 1.1.
1.2.2 Verify units: For Python sources, we review that the unit management Pint library has been applied correctly. For spreadsheets we ensure units are consistent and clearly marked in the spreadsheet. 
1.2.3 Label and Format: Ensure the source has clear labels, naming, comments and consistent formatting. This is to ensure the source calculation is easy to understand for users.
1.2.4 Annotations: For complex calculations, we’ll add more detailed annotations and explanations as per 1.1 that help users understand the methodology and assumptions behind the calculations. 
﻿
1.3 Compare Results with Existing SoftwareTo ensure our calculations are in line with competitor products, as part of our review process, we directly compare our calculation results with multiple existing worked examples. Differences identified in results allow us to identify any shortcomings of our calculations and correct them. Additionally we take this opportunity to ensure CalcTree calculations offer more robust and in-depth checks compared to existing literature.
1.3.1 Select worked examples: CalcTree uses industry-recognised engineering solutions as a benchmark for each specific calculation type. This ensures that our methods align with accepted standards.
1.3.2 Perform Calculations and compare results: We execute calculations using both our calculator and the selected benchmark example to generate results for comparison. We then compare the results and examine any differences and/or similarities.
1.3.3 Address Discrepancies: Any discrepancies between CalcTree’s calculations and the comparison benchmark are identified, allowing us to pinpoint potential errors in assumptions, input data, or formulas. We then make precise adjustments to calculations to correct any discrepancies. This may involve refining formulas, correcting data, or revisiting assumptions to enhance accuracy.
﻿
2 - Page Content reviewOnce we’ve confirmed the correctness and optimised the source calculation, stage 2 moves to verifying the accompanying page content.
2.1 Parameters, units and formulas2.1.1 Parameter names and descriptions: Ensure parameters are named correctly and in alignment with the calculation source if applicable. Also add to and refine parameter descriptions where appropriate.
2.1.2 Parameter linking: Ensure all page parameters, tables and charts are linked to the correct component in the calculation source.
2.1.3 Parameter formulas: For parameters performing additional operations outside of the source, we ensure the formula is correct and linked up correctly. 
2.1.4 Ensure units and parameter names are consistent: Double check parameter units are correct and linked to the source appropriately in the case of Python calculation sources. Otherwise the set units in the parameters are aligned with spreadsheet source units.
2.2 Clarity2.2.1 Ensure Clear Communication: CalcTree prioritises clarity in all content, including instructions, explanations, and descriptions. We use straightforward language and avoid unnecessary jargon to cater to a diverse audience, ensuring that everyone can understand our calculations easily.
2.2.2 Supporting Diagrams and Images: Visual aids such as diagrams and images enhance understanding and provide context. CalcTree ensures that all visuals are relevant, high-quality, and accurately represent the calculations and concepts, making complex information more accessible.
2.2.3 Ensure Code References Are Clear: For calculations referencing design codes, we ensure that references are correct and visible on the page, suitable for a printed report. We highlight code snippets and provide explanations to help users understand their purpose impact on calculation results.
﻿
3 Performance ReviewPage performance plays a vital role in user experience. Stage 3 reviews the general performance of the calculator and seeks to optimise any user experience errors or issues that might exist.
3.1 Calculation SpeedA key factor in user satisfaction is the speed at which calculations are performed. This step focuses on optimising algorithms and processes to deliver quick and accurate results, ensuring users can rely on CalcTree for efficient performance.
3.1.1 Calculation robustness: This step checks that each individual calculation/parameter in the overall calculator works. Calculations that break are first reviewed by the reviewer to identify any issues with the calculation source or otherwise. Persistent errors are shared with our software engineering team for further review and resolution.
3.1.2 Calculation Speed: This step checks that the speed of each calculation in the overall calculator works at a reasonable pace. Calculations identified as slow are first reviewed by the reviewer, to identify any issues with the calculation source or otherwise. Persistent speed issues are shared with our software engineering team for further review and resolution.
3.1.3 Apply improvements: Where possible improvements to the calculation source, or page impacting speed and robustness are applied immediately by the reviewer. In cases where an underlying bug with the platform causing issues has been identified, the issue enters our software engineering bug resolution process, which sits at the top of our team's priority list. In some cases, this step requires returning to stage 1.2 to optimise the source calculation. 
﻿
3.2 Page Load SpeedSimilarly to the individual parameters and calculations on our pages, we review that the overall page is snappy and loads quickly to ensure a positive user experience.
3.2.1 Ensure Fast Page Loading: Using varying global IP addresses and relevant monitoring tools (DataDog & Google Analytics) we measure the load speed of the page. ‘Heavy’ items such as videos, gifs and large images can slow page load times. If the load speed is outside our benchmarks, we seek to reduce the weight of content on the page where possible.
3.2.2 Monitor and Improve: CalcTree regularly assesses page load speeds more broadly as well. If a calculator is identified as slow in these regular checks, our team jumps into this process to resolve it.
ConclusionBuilding confidence in engineering calculations means covering all the bases: technical accuracy, clear organisation, solid page performance, and straightforward communication. By following this checklist, we build strong confidence in the tools we’re providing to the market. 
﻿
That said, we’re always looking to improve, and your feedback plays a big part in that. We’d love to hear what you think about our platform. Got an idea for a new feature? Spotted something that could be better? Reach out via our in-app chat or join our Slack community. Your insights help us make CalcTree even better for everyone.
﻿