The popularization of programming languages, such as Python and R, offers a paradigm shift in how companies can calculate their most important indicators. The flexibility of these languages offers a wide range of functions and capabilities, allowing analysts to customize code to meet their company's specific needs.
In recent years, the massive amount of data generated through Big Data has been a challenge for many organizations, in which many still use traditional spreadsheets to manage and calculate their logistical indicators, which naturally involve a large amount of data. However, as the volume of data generated increases over time, the capacity of such spreadsheets has become less and less sufficient. A database with more than 1 million rows, which is not uncommon in daily stock position databases, for example, becomes unfeasible to work with in these spreadsheets.
In contrast, programming languages are capable of handling massive amounts of data. In addition, there is still the possibility of processing the code on servers in the cloud, which eliminates the need to use computers with very high processing power. Dealing “by hand” with such an amount of data is very prone to human error, in addition to being time consuming. Therefore, programming languages offer the advantage of reliability, as they are capable of performing complex calculations with precision, avoiding human errors and reducing the time needed to perform the analyses.
Another advantage of programming languages is automation, as once the code is written, the process of updating indicators becomes much more automated. Changes and updates in databases with new observations, as long as they maintain the same structure as the original base, only imply running the code again to generate updated results. This brings benefits mainly in the calculation of indicators whose update frequency is high, such as daily, weekly and monthly indicators. This automation optimizes the employee's time, who can focus more time on other activities.
Programming languages also have the advantage of having an active community that frequently releases new libraries, in addition to updating existing ones, which have several features that can be easily integrated into the code. Libraries like pandas (for Python), and the dplyr (for R), were developed to facilitate data analysis and mathematical calculations. There are also libraries that allow reading and saving files in different formats, such as csv, xlsx, among others. Thus, it is possible to generate results in files with more suitable formats to be used in other programs and applications, such as Excel and Power BI, for example.
Other interesting libraries for logistic indicators are the plotly (for Python) and ggplot2 (for R), focused on creating different types of graphs. There is also the library geobr, (for both languages) developed to facilitate the creation of different types of maps with geographic data from Brazil. Finally, there are also libraries focused on collecting data from the internet, a practice known as web scraping, which provides the advantage of collecting data online without the need for a manual process to do so.
At ILOS, we use programming languages for data analysis in projects and in calculations and updates of internal indicators, which boosts our team's productivity and speeds up the delivery of results to clients. To illustrate what can be done with such languages, below are examples of indicators and maps generated through public data collected online, referring to the number of storage facilities for agricultural products across the country, as well as their capacity in tons, in the first half of 2022.
Source: IBGE, Directorate of Research, Coordination of Agricultural Statistics, Inventory Research, 1st half of 2022. ILOS analysis
In summary, the use of programming languages in the calculation and monitoring of logistic indicators offers a series of advantages, from flexibility in customizing codes to reliability, analysis capacity, automation and the possibility of collecting data through web scraping. With these tools, companies can obtain valuable insights into their logistics performance and make more agile and qualified decisions to monitor their results and improve their processes.
References:
- https://agenciadenoticias.ibge.gov.br/agencia-sala-de-imprensa/2013-agencia-de-noticias/releases/35416-capacidade-de-armazenagem-agricola-cresce-3-0-e-chega-a-188-8-milhoes-de-toneladas-no-1-semestre-de-2022
- https://www.ibge.gov.br/estatisticas/downloads-estatisticas.html
- https://pandas.pydata.org/
- https://dplyr.tidyverse.org/
- https://plotly.com/python/
- https://ggplot2.tidyverse.org/
- https://www.urbandemographics.org/software/
