Data visualization is a fundamental practice to democratize access to public information. However, creating effective graphics goes far beyond choosing attractive colors or using the latest technological tools. As Alberto Cairo, an expert in data visualization and professor at the academy of the European Open Data Portal (data.europa.eu), points out, "every design decision must be deliberate: inevitably subjective, but never arbitrary." Through a series of three webinars that you can watch again here, the expert offered innovative tips to be at the forefront of data visualization.

When working with data visualization, especially in the context of public information, it is crucial to debunk some myths ingrained in our professional culture. Phrases like "data speaks for itself," "a picture is worth a thousand words," or "show, don't count" sound good, but they hide an uncomfortable truth: charts don't always communicate automatically.

The reality is more complex. A design professional may want to communicate something specific, but readers may interpret something completely different. How can you bridge the gap between intent and perception in data visualization? In this post, we offer some keys to the training series.

A structured framework for designing with purpose

Rather than following rigid "rules" or applying predefined templates, the course proposes a framework of thinking based on five interrelated components:

1. Content: the nature, origin, and limitations of the data
2. People: The audience we are targeting
3. Intention: The Purposes We Define
4. Constraints: The Constraints We Face
5. Results: how the graph is received

This holistic approach forces us to constantly ask ourselves: what do our readers really need to know? For example, when communicating information about hurricane or health emergency risks, is it more important to show exact trajectories or communicate potential impacts? The correct answer depends on the context and, above all, on the information needs of citizens.

The danger of over-aggregation

Even without losing sight of the purpose, it is important not to fall into adding too much information or presenting only averages. Imagine, for example, a dataset on citizen security at the national level: an average may hide the fact that most localities are very safe, while a few with extremely high rates distort the national indicator.

As Claus O. Wilke explains in his book "Fundamentals of Data Visualization," this practice can hide crucial patterns, outliers, and paradoxes that are precisely the most relevant to decision-making. To avoid this risk, the training proposes to visualize a graph as a system of layers that we must carefully build from the base:

1. Encoding

1. It's the foundation of everything: how we translate data into visual attributes. Research in visual perception shows us that not all "visual channels" are equally effective. The hierarchy would be:

• Most effective: position, length and height
• Moderately effective: angle, area and slope
• Less effective: color, saturation, and shape

How do we put this into practice? For example, for accurate comparisons, a bar chart will almost always be a better choice than a pie chart. However, as nuanced in the training materials, "effective" does not always mean "appropriate". A pie chart can be perfect when we want to express the idea of a "whole and its parts", even if accurate comparisons are more difficult.

2. Arrangement 

1. The positioning, ordering, and grouping of elements profoundly affects perception. Do we want the reader to compare between categories within a group, or between groups? The answer will determine whether we organize our visualization with grouped or stacked bars, with multiple panels, or in a single integrated view.

3. Scaffolding

Titles, introductions, annotations, scales and legends are fundamental. In datos.gob.es we've seen how interactive visualizations can condense complex information, but without proper scaffolding, interactivity can confuse rather than clarify.

The value of a correct scale

One of the most delicate – and often most manipulable – technical aspects of a visualization is the choice of scale. A simple modification in the Y-axis can completely change the reader's interpretation: a mild trend may seem like a sudden crisis, or sustained growth may go unnoticed.

As mentioned in the second webinar in the series, scales are not a minor detail: they are a narrative component. Deciding where an axis begins, what intervals are used, or how time periods are represented involves making choices that directly affect one's perception of reality. For example, if an employment graph starts the Y-axis at 90% instead of 0%, the decline may seem dramatic, even if it's actually minimal.

Therefore, scales must be honest with the data. Being "honest" doesn't mean giving up on design decisions, but  rather clearly showing what decisions were made and why. If there is a valid reason for starting the Y-axis at a non-zero value, it should be explicitly explained in the graph or in its footnote. Transparency must prevail over drama.

Visual integrity not only protects the reader from misleading interpretations, but also reinforces the credibility of the communicator. In the field of public data, this honesty is not optional: it is an ethical commitment to the truth and to citizen trust.

Accessibility: Visualize for everyone

On the other hand, one of the aspects often forgotten is accessibility. About 8% of men and 0.5% of women have some form of color blindness. Tools like Color Oracle allow you to simulate what our visualizations look like for people with different types of color perception impairments.

In addition, the webinar mentioned the Chartability project, a methodology to evaluate the accessibility of data visualizations. In the Spanish public sector, where web accessibility is a legal requirement, this is not optional: it is a democratic obligation. Under this premise, the Spanish Federation of Municipalities and Provinces published a Data Visualization Guide for Local Entities. 

Visual Storytelling: When Data Tells Stories

Once the technical issues have been resolved, we can address the narrative aspect that is increasingly important to communicate correctly. In this sense, the course proposes a simple but powerful method:

1. Write a long sentence that summarizes the points you want to communicate.
2. Break that phrase down into components, taking advantage of natural pauses.
3. Transform those components into sections of your infographic.

This narrative approach is especially effective for projects like the ones we found in data.europa.eu, where visualizations are combined with contextual explanations to communicate the value of high-value datasets or in datos.gob.es's data science and visualization exercises.

The future of data visualization also includes more creative and user-centric approaches. Projects that incorporate personalized elements, that allow readers to place themselves at the center of information, or that use narrative techniques to generate empathy, are redefining what we understand by "data communication".

Alternative forms of "data sensification" are even emerging: physicalization (creating three-dimensional objects with data) and sonification (translating data into sound) open up new possibilities for making information more tangible and accessible. The Spanish company Tangible Data, which we echo in datos.gob.es because it reuses open datasets, is proof of this.

Figure 1. Examples of data sensification. Source: https://data.europa.eu/sites/default/files/course/webinar-data-visualisation-episode-3-slides.pdf

By way of conclusion, we can emphasize that integrity in design is not a luxury: it is an ethical requirement. Every graph we publish on official platforms influences how citizens perceive reality and make decisions. That is why mastering technical tools such as libraries and visualization APIs, which are discussed in other articles on the portal, is so relevant.

The next time you create a visualization with open data, don't just ask yourself "what tool do I use?" or "Which graphic looks best?". Ask yourself: what does my audience really need to know? Does this visualization respect data integrity? Is it accessible to everyone? The answers to these questions are what transform a beautiful graphic into a truly effective communication tool.

Imagine you want to know how many terraces there are in your neighbourhood, how the pollen levels in the air you breathe every day are evolving or whether recycling in your city is working well. All this information exists in your municipality's databases, but it sits in spreadsheets and technical documents that only experts know how to interpret.

This is where open data visualisation initiativescome in: they transform those seemingly cold numbers into stories that anyone can understand at a glance.  A colourful graph showing the evolution of traffic on your street, an interactive map showing the green areas of your city, or an infographic explaining how the municipal budget is spent. These tools make public information accessible, useful and, moreover, comprehensible to all citizens.

Moreover, the advantages of this type of solution are not only for the citizens, but also benefit the Administration that carries out the exercise, because it allows:

• Detect and correct data errors.
• Add new sets to the portal.
• Reduce the number of questions from citizens.
• Generate more trust on the part of society.

Therefore, visualising open data brings government closer to citizens, facilitates informed decision-making, helps public administrations to improve their open data offer and creates a more participatory society where we can all better understand how the public sector works. In this post, we present some examples of open data visualisation initiatives in regional and municipal open data portals.

Visualiza Madrid: bringing data closer to the public

Madrid City Council's open data portal has developed the initiative "Visualiza Madrid", a project born with the specific objective of making open data and its potential reach the general public , transcending specialised technical profiles. As Ascensión Hidalgo Bellota, Deputy Director General for Transparency of Madrid City Council, explained during the IV National Meeting on Open Data, "this initiative responds to the need to democratise access to public information".

Visualiza Madrid currently has 29 visualisations that cover different topics of interest to citizens, from information on hotel and restaurant terraces to waste management and urban traffic analysis. This thematic diversity demonstrates the versatility of visualisations as a tool for communicating information from very diverse sectors of public administration.

 In addition, the initiative has received external recognition this year through the Audaz 2,025 Awards, an initiative of the Spanish chapter of the Open Government Academic Network (RAGA Spain).The initiative has also received external recognition through the Audaz 2,025 Awards.

Castilla y León: comprehensive analysis of regional data

 The Junta de Castilla y León has also developed a portal specialised in analysis and visualisations that stands out for its comprehensive approach to the presentation of regional data. Its visualisation platform offers a systematic approach to the analysis of regional information, allowing users to explore different dimensions of the reality of Castilla y Leónthrough interactive and dynamic tools.

This initiative allows complex information to be presented in a structured and understandable way, facilitating both academic analysis and citizen use of the data. The platform integrates different sources of regional information, creating a coherent ecosystem of visualisations that provides a panoramic view of different aspects of regional management. Among the topics it offers are data on tourism, the labour market and budget execution. All the visualisations are made with open data sets from the regional portal of Castilla y León .

The Castilla y León approach demonstrates how visualisations can serve as a tool for territorial analysis, providing valuable insights on economic, social and demographic dynamics that are fundamental for the planning and evaluation of regional public policies.

Canary Islands: technological integration with interactive widgets .

On the other hand, the Government of the Canary Islands has opted for an innovative strategy through the implementation of widgets that allow the integration of open data visualisations of the Instituto Canario de Estadística (ISTAC) in different platforms and contexts. This technological approach represents a qualitative leap in the distribution and reuse of public data visualisations.

The widgets developed by the Canary Islands make it easier for third parties to embed official visualisations in their own applications, websites or analyses, exponentially expanding the scope and usefulness of Canary Islands open data. This strategy not only multiplies the points of access to public information, but also fosters the creation of a collaborative ecosystem where different actors can benefit from and contribute to the value of open data.

The Canarian initiative illustrates how technology can be used to create scalable and flexible solutions that maximise the impact of investments in open data visualisation, establishing a replicable model for other administrations seeking to amplify the reach of their transparency initiatives.

Lessons learned and best practices

By way of example, the cases analysed reveal common patterns that can serve as a guide for future initiatives. The orientation towards the general public, beyond specialised technical users, emerges as an opportunity factor for the success of these platforms. To maintain the interest and relevance of the visualisations, it is important to offer thematic diversity and to update the data regularly.

Technological integration and interoperability, as demonstrated in the case of the Canary Islands, open up new possibilities to maximise the impact of public investments in data visualisation. Likewise, external recognition and participation in professional networks, as evidenced in the case of Madrid, contribute to continuous improvement and the exchange of best practices between administrations.

In general terms, open data visualisation initiatives represent a very valuable opportunity in the transparency and open government strategy of Spanish public administrations. The cases of Madrid, Castilla y León, as well as the Canary Islands, are examples of the enormous potential for transforming public data into tools for citizen empowerment and improved public management.

The success of these initiatives lies in their ability to connect government information with the real needs of citizens, creating bridges of understanding that strengthen the relationship between administration and society. As these experiences mature and consolidate, it will be crucial to keep the focus on the usability, accessibility and relevance of visualisations, ensuring that open data truly delivers on its promise to contribute to a more informed, participatory and democratic society.

Open data visualisation is not just a technical issue, but a strategic opportunity to redefine public communication and strengthen the foundations of a truly open and transparent administration.

The European Drug Report provides a current overview of the drug situation in the region, analysing the main trends and emerging threats. It is a valuable publication, with a high number of downloads, which is quoted in many media outlets.

The report is produced annually by the European Union Drugs Agency (EUDA), the current name of the former European Monitoring Centre for Drugs and Drug Addiction. It collects and analyses data from EU Member States, together with other partner countries such as Turkey and Norway, to provide a comprehensive picture of drug use and supply, drug harms and harm reduction interventions. The report contains comprehensive datasets on these issues disaggregated at the national level, and even, in some cases, at the city level (such as Barcelona or Palma de Mallorca).

This study has been carried out since 1993 and translated into more than 20 official languages of the European Union. However, in the last two years it has introduced a new feature: a change in internal processes to improve the visualisation of the data obtained. A process they explained in the recent webinar "The European Drug Report: using an open data approach to improve data visualisation", organised by the European Open Data Portal (data.europa.eu) on 25 June. The following is a summary of what the Observatory's representatives had to say at this event.

The need for change

The Observatory has always worked with open data, but there were inefficiencies in the process. Until now, the European Drug Report has always been published in PDF format, with the focus on achieving a visually appealing product. The internal process leading up to the publication of the report consisted of several stages involving various teams: 

1. A team from the Observatory checked the format of the data received from the supplier and, if necessary, adapted it.
2. A specialised data analysis team created visualisations from the data.
3. A specialised drafting team drafted the report. The team that had created the visualisations could collaborate in this phase.
4. An internal team validated the content of the report.
5. The data provider checked that the Observatory had interpreted the data correctly.

Despite the good reception of the report and its format, in 2022 the Observatory decided to completely change the publication format for the following reasons:

• Once the various steps of the publication process had been initiated, the data were formatted and were no longer machine-readable. This reduced the accessibility of the data, e.g. for screen readers, and limited its reusability.
• If errors were detected in the different steps of the process, they were corrected directly on the format of the data in this step. In other words, if an error was detected in a chart during the revision phase, it was corrected directly on that chart. This procedure could cause errors and dull the traceability of data, limiting efficiency: the same static graph could be present several times in the document and each mention had to be corrected individually.
• At the end of the process, the format of the source data had to be adjusted due to changes in the publication procedure.
• Many of the users who consulted the report did so from a mobile device, for which the PDF format was not always suitable.
• Because they are neither accessible nor mobile-friendly, PDF documents did not usually appear as the first result in search engines. This point is important for the Observatory, as many users find the report through search engines.

A responsive web format was needed, which automatically adjusts a website to the size and layout of its users' devices.  The aim was to:

• Improved accessibility.
• A more streamlined process for creating visualisations.
• An easier translation process.
• An increase in visitors from search engines.
• Greater modularity.

The process behind the new report

In order to completely transform the publication format of the report, an ad hoc visualisation process has been carried out, summarised in the following image:

Figure 1. Process for creating visualizations for the European Drug Report. Source EN: Webinar “The European Drug Report using an open data approach to improve data visualisation”, organized by data.europa.eu.

The main new feature is that visualisations are created dynamically from the source data. In this way, if something is changed in these data, it is automatically changed in all visualisations that feed on it. Using the Drupal content management system, on which much of the site is based, administrators can register changes that will automatically be reflected in the HTML and therefore in the displays. In addition, site administrators have a visualisation generator which, based on data and indications - equivalent to simple instructions such as "sort from highest to lowest" expressed in HTML - creates visualisations without the need to touch code.

The same dynamic update procedure applies to the PDF that the user can download. If there are changes in the data, in the visualisations or if typographical errors are corrected, the PDF is generated again through a compilation process that the Observatory has created specifically for this task.

The report after the change

The report is currently published in HTML version, with the possibility to download chapters or the full report in PDF format. It is structured by thematic modules and also allows the consultation of annexes.

Furthermore, the data are always published in CSV format and the licensing conditions of the data (CC-BY-4.0) are indicated on the same page. The reference to the source of the data is always made available to the reader on the same page as a visualisation.

With this change in procedure and format, benefits for all have been achieved. From the readers' point of view, the user experience has been improved. For the organisation, the publication process has been streamlined.

In terms of open data, this new approach allows for greater traceability, as the data can be consulted at any time in its current format. Moreover, according to the Observatory speakers, this new format of the report, together with the fact that the data and visualisations are always up-to-date, has increased the accessibility of the data for the media.

You can access the webinar materials here:

• Slides
• Recording

1. Introduction

In the information age, artificial intelligence has proven to be an invaluable tool for a variety of applications. One of the most incredible manifestations of this technology is GPT (Generative Pre-trained Transformer), developed by OpenAI. GPT is a natural language model that can understand and generate text, providing coherent and contextually relevant responses. With the recent introduction of Chat GPT-4, the capabilities of this model have been further expanded, allowing for greater customisation and adaptability to different themes.

In this post, we will show you how to set up and customise a specialised critical minerals wizard using GPT-4 and open data sources. As we have shown in previous publications critical minerals are fundamental to numerous industries, including technology, energy and defence, due to their unique properties and strategic importance. However, information on these materials can be complex and scattered, making a specialised assistant particularly useful.

The aim of this post is to guide you step by step from the initial configuration to the implementation of a GPT wizard that can help you to solve doubts and provide valuable information about critical minerals in your day to day life. In addition, we will explore how to customise aspects of the assistant, such as the tone and style of responses, to perfectly suit your needs. At the end of this journey, you will have a powerful, customised tool that will transform the way you access and use critical open mineral information.

Access the data lab repository on Github.

2. Context

The transition to a sustainable future involves not only changes in energy sources, but also in the material resources we use. The success of sectors such as energy storage batteries, wind turbines, solar panels, electrolysers, drones, robots, data transmission networks, electronic devices or space satellites depends heavily on access to the raw materials critical to their development. We understand that a mineral is critical when the following factors are met:

• Its global reserves are scarce
• There are no alternative materials that can perform their function (their properties are unique or very unique)
• They are indispensable materials for key economic sectors of the future, and/or their supply chain is high risk

You can learn more about critical minerals in the post mentioned above.

3. Target

This exercise focuses on showing the reader how to customise a specialised GPT model for a specific use case. We will adopt a "learning-by-doing" approach, so that the reader can understand how to set up and adjust the model to solve a real and relevant problem, such as critical mineral expert advice. This hands-on approach not only improves understanding of language model customisation techniques, but also prepares readers to apply this knowledge to real-world problem solving, providing a rich learning experience directly applicable to their own projects.

The GPT assistant specialised in critical minerals will be designed to become an essential tool for professionals, researchers and students. Its main objective will be to facilitate access to accurate and up-to-date information on these materials, to support strategic decision-making and to promote education in this field. The following are the specific objectives we seek to achieve with this assistant:

• Provide accurate and up-to-date information:
  
  • The assistant should provide detailed and accurate information on various critical minerals, including their composition, properties, industrial uses and availability.
  • Keep up to date with the latest research and market trends in the field of critical minerals.
• Assist in decision-making:
  
  • To provide data and analysis that can assist strategic decision making in industry and critical minerals research.
  • Provide comparisons and evaluations of different minerals in terms of performance, cost and availability.
• Promote education and awareness of the issue:
  
  • Act as an educational tool for students, researchers and practitioners, helping to improve their knowledge of critical minerals.
  • Raise awareness of the importance of these materials and the challenges related to their supply and sustainability.

4. Resources

To configure and customise our GPT wizard specialising in critical minerals, it is essential to have a number of resources to facilitate implementation and ensure the accuracy and relevance of the model''s responses. In this section, we will detail the necessary resources that include both the technological tools and the sources of information that will be integrated into the assistant''s knowledge base.

Tools and Technologies

The key tools and technologies to develop this exercise are:

• OpenAI account: required to access the platform and use the GPT-4 model. In this post, we will use ChatGPT''s Plus subscription to show you how to create and publish a custom GPT. However, you can develop this exercise in a similar way by using a free OpenAI account and performing the same set of instructions through a standard ChatGPT conversation.
• Microsoft Excel: we have designed this exercise so that anyone without technical knowledge can work through it from start to finish. We will only use office tools such as Microsoft Excel to make some adjustments to the downloaded data.

In a complementary way, we will use another set of tools that will allow us to automate some actions without their use being strictly necessary:

• Google Colab: is a Python Notebooks environment that runs in the cloud, allowing users to write and run Python code directly in the browser. Google Colab is particularly useful for machine learning, data analysis and experimentation with language models, offering free access to powerful computational resources and facilitating collaboration and project sharing.
• Markmap: is a tool that visualises Markdown mind maps in real time. Users write ideas in Markdown and the tool renders them as an interactive mind map in the browser. Markmap is useful for project planning, note taking and organising complex information visually. It facilitates understanding and the exchange of ideas in teams and presentations.

Sources of information

• Raw Materials Information System (RMIS): raw materials information system maintained by the Joint Research Center of the European Union. It provides detailed and up-to-date data on the availability, production and consumption of raw materials in Europe.
• International Energy Agency (IEA) Catalogue of Reports and Data: the International Energy Agency (IEA) offers a comprehensive catalogue of energy-related reports and data, including statistics on production, consumption and reserves of energy and critical minerals.
• Mineral Database of the Spanish Geological and Mining Institute (BDMIN in its acronym in Spanish): contains detailed information on minerals and mineral deposits in Spain, useful to obtain specific data on the production and reserves of critical minerals in the country.

With these resources, you will be well equipped to develop a specialised GPT assistant that can provide accurate and relevant answers on critical minerals, facilitating informed decision-making in the field.

5. Development of the exercise

5.1. Building the knowledge base

For our specialised critical minerals GPT assistant to be truly useful and accurate, it is essential to build a solid and structured knowledge base. This knowledge base will be the set of data and information that the assistant will use to answer queries. The quality and relevance of this information will determine the effectiveness of the assistant in providing accurate and useful answers.

We start with the collection of information sources that will feed our knowledge base. Not all sources of information are equally reliable. It is essential to assess the quality of the sources identified, ensuring that:

• Information is up to date: the relevance of data can change rapidly, especially in dynamic fields such as critical minerals.
• The source is reliable and recognised: it is necessary to use sources from recognised and respected academic and professional institutions.
• Data is complete and accessible: it is crucial that data is detailed and accessible for integration into our wizard.

 In our case, we developed an online search in different platforms and information repositories trying to select information belonging to different recognised entities:

• Research centres and universities:
  
  • They publish detailed studies and reports on the research and development of critical minerals.
  • Example: RMIS of the Joint Research Center of the European Union.
• Governmental institutions and international organisations:
  
  • These entities usually provide comprehensive and up-to-date data on the availability and use of critical minerals.
  • Example: International Energy Agency (IEA).
• Specialised databases:
  
  • They contain technical and specific data on deposits and production of critical minerals.
  • Example: Minerals Database of the Spanish Geological and Mining Institute (BDMIN).

Selection and preparation of information

We will now focus on the selection and preparation of existing information from these sources to ensure that our GPT assistant can access accurate and useful data.

RMIS of the Joint Research Center of the European Union:

• Selected information:

We selected the report "Supply chain analysis and material demand forecast in strategic technologies and sectors in the EU - A foresight study". This is an analysis of the supply chain and demand for minerals in strategic technologies and sectors in the EU. It presents a detailed study of the supply chains of critical raw materials and forecasts the demand for minerals up to 2050.

• Necessary preparation: 

The format of the document, PDF, allows the direct ingestion of the information by our assistant. However, as can be seen in Figure 1, there is a particularly relevant table on pages 238-240 which analyses, for each mineral, its supply risk, typology (strategic, critical or non-critical) and the key technologies that employ it. We therefore decided to extract this table into a structured format (CSV), so that we have two pieces of information that will become part of our knowledge base.

Figure 1: Table of minerals contained in the JRC PDF

To programmatically extract the data contained in this table and transform it into a more easily processable format, such as CSV(comma separated values), we will use a Python script that we can use through the platform Google Colab platform (Figure 2).

Figure 2: Script Python para la extracción de datos del PDF de JRC desarrollado en plataforma Google Colab.

To summarise, this script:

1. It is based on the open source library PyPDF2capable of interpreting information contained in PDF files.
2. First, it extracts in text format (string) the content of the pages of the PDF where the mineral table is located, removing all the content that does not correspond to the table itself.
3. It then goes through the string line by line, converting the values into columns of a data table. We will know that a mineral is used in a key technology if in the corresponding column of that mineral we find a number 1 (otherwise it will contain a 0).
4. Finally, it exports the table to a CSV file for further use.

International Energy Agency (IEA):

• Selected information:

We selected the report "Global Critical Minerals Outlook 2024". It provides an overview of industrial developments in 2023 and early 2024, and offers medium- and long-term prospects for the demand and supply of key minerals for the energy transition. It also assesses risks to the reliability, sustainability and diversity of critical mineral supply chains.

• Necessary preparation:

The format of the document, PDF, allows us to ingest the information directly by our virtual assistant. In this case, we will not make any adjustments to the selected information.

Spanish Geological and Mining Institute''s Minerals Database (BDMIN)

• Selected information:

In this case, we use the form to select the existing data in this database for indications and deposits in the field of metallogeny, in particular those with lithium content.

Figure 3: Dataset selection in BDMIN.

• Necessary preparation:

We note how the web tool allows online visualisation and also the export of this data in various formats. Select all the data to be exported and click on this option to download an Excel file with the desired information.

Figure 4: Visualization and download tool in BDMIN

Figure 5: BDMIN Downloaded Data.

All the files that make up our knowledge base can be found at GitHub, so that the reader can skip the downloading and preparation phase of the information.

5.2. GPT configuration and customisation for critical minerals

When we talk about "creating a GPT," we are actually referring to the configuration and customisation of a GPT (Generative Pre-trained Transformer) based language model to suit a specific use case. In this context, we are not creating the model from scratch, but adjusting how the pre-existing model (such as OpenAI''s GPT-4) interacts and responds within a specific domain, in this case, on critical minerals.

First of all, we access the application through our browser and, if we do not have an account, we follow the registration and login process on the ChatGPT platform. As mentioned above, in order to create a GPT step-by-step, you will need to have a Plus account. However, readers who do not have such an account can work with a free account by interacting with ChatGPT through a standard conversation.

Figure 6: ChatGPT login and registration page.

Once logged in, select the "Explore GPT" option, and then click on "Create" to begin the process of creating your GPT.

Figure 7: Creation of new GPT.

The screen will display the split screen for creating a new GPT: on the left, we will be able to talk to the system to indicate the characteristics that our GPT should have, while on the left we will be able to interact with our GPT to validate that its behaviour is adequate as we go through the configuration process.

Figure 8: Screen of creating new GPT.

In the GitHub of this project, we can find all the prompts or instructions that we will use to configure and customise our GPT and that we will have to introduce sequentially in the "Create" tab, located on the left tab of our screens, to complete the steps detailed below.

The steps we will follow for the creation of the GPT are as follows:

1. First, we will outline the purpose and basic considerations for our GPT so that you can understand how to use it.

Figure 9: Basic instructions for new GPT.

2. We will then create a name and an image to represent our GPT and make it easily identifiable. In our case, we will call it MateriaGuru.

Figure 10: Name selection for new GPT.

Figure 11: Image creation for GPT.

3.We will then build the knowledge base from the information previously selected and prepared to feed the knowledge of our GPT.

Figure 12: Uploading of information to the new GPT knowledge base.

4. Now, we can customise conversational aspects such as their tone, the level of technical complexity of their response or whether we expect brief or elaborate answers.

5. Lastly, from the "Configure" tab, we can indicate the  conversation starters desired so that users interacting with our GPT have some ideas to start the conversation in a predefined way.

Figure 13: Configure GPT tab.

In Figure 13 we can also see the final result of our training, where key elements such as their image, name, instructions, conversation starters or documents that are part of their knowledge base appear.

5.3. Validation and publication of GPT

Before we sign off our new GPT-based assistant, we will proceed with a brief validation of its correct configuration and learning with respect to the subject matter around which we have trained it. For this purpose, we prepared a battery of questions that we will ask MateriaGuru to check that it responds appropriately to a real scenario of use.

#	Question	Answer
1	Which critical minerals have experienced a significant drop in prices in 2023?	Battery mineral prices saw particularly large drops with lithium prices falling by 75% and cobalt, nickel and graphite prices falling by between 30% and 45%.
2	What percentage of global solar photovoltaic (PV) capacity was added by China in 2023?	China accounted for 62% of the increase in global solar PV capacity in 2023.
3	What is the scenario that projects electric car (EV) sales to reach 65% by 2030?	The Net Zero Emissions (NZE) scenario for 2050 projects that electric car sales will reach 65% by 2030.
4	What was the growth in lithium demand in 2023?	Lithium demand increased by 30% in 2023.
5	Which country was the largest electric car market in 2023?	China was the largest electric car market in 2023 with 8.1 million electric car sales representing 60% of the global total.
6	What is the main risk associated with market concentration in the battery graphite supply chain?	More than 90% of battery-grade graphite and 77% of refined rare earths in 2030 originate in China, posing a significant risk to market concentration.
7	What proportion of global battery cell production capacity was in China in 2023?	China owned 85% of battery cell production capacity in 2023.
8	How much did investment in critical minerals mining increase in 2023?	Investment in critical minerals mining grew by 10% in 2023.
9	What percentage of battery storage capacity in 2023 was composed of lithium iron phosphate (LFP) batteries?	By 2023, LFP batteries would constitute approximately 80% of the total battery storage market.
10	What is the forecast for copper demand in a net zero emissions (NZE) scenario for 2040?	In the net zero emissions (NZE) scenario for 2040, copper demand is expected to have the largest increase in terms of production volume.

Figure 14: Table with battery of questions for the validation of our GPT.

Using the preview section on the right-hand side of our screens, we launch the battery of questions and validate that the answers correspond to those expected.

Figure 15: Validation of GPT responses.

Finally, click on the "Create" button to finalise the process. We will be able to select between different alternatives to restrict its use by other users.

Figure 16: Publication of our GPT.

6. Scenarios of use

In this section we show several scenarios in which we can take advantage of MateriaGuru in our daily life. On the GitHub of the project you can find the prompts used to replicate each of them.

6.1. Consultation of critical minerals information

The most typical scenario for the use of this type of GPTs is assistance in resolving doubts related to the topic in question, in this case, critical minerals. As an example, we have prepared a set of questions that the reader can pose to the GPT created to understand in more detail the relevance and current status of a critical material such as graphite from the reports provided to our GPT.

 

Figure 17: Resolution of critical mineral queries.

We can also ask you specific questions about the tabulated information provided on existing sites and evidence on Spanish territory.

Figure 18: Lithium reserves in Extremadura.

6.2. Representation of quantitative data visualisations

Another common scenario is the need to consult quantitative information and make visual representations for better understanding. In this scenario, we can see how MateriaGuru is able to generate an interactive visualisation of graphite production in tonnes for the main producing countries.

Figure 19: Interactive visualisation generation with our GPT.

6.3. Generating mind maps to facilitate understanding

Finally, in line with the search for alternatives for a better access and understanding of the existing knowledge in our GPT, we will propose to MateriaGuru the construction of a mind map that allows us to understand in a visual way key concepts of critical minerals. For this purpose, we use the open Markmap notation (Markdown Mindmap), which allows us to define mind maps using markdown notation.

Figure 20: Generation of mind maps from our GPT

We will need to copy the generated code and enter it in a  markmapviewer in order to generate the desired mind map. We facilitate here a version of this code generated by MateriaGuru.

Figure 21: Visualisation of mind maps.

7. Results and conclusions

In the exercise of building an expert assistant using GPT-4, we have succeeded in creating a specialised model for critical minerals. This wizard provides detailed and up-to-date information on critical minerals, supporting strategic decision making and promoting education in this field. We first gathered information from reliable sources such as the RMIS, the International Energy Agency (IEA), and the Spanish Geological and Mining Institute (BDMIN). We then process and structure the data appropriately for integration into the model. Validations showed that the wizard accurately answers domain-relevant questions, facilitating access to your information.

In this way, the development of the specialised critical minerals assistant has proven to be an effective solution for centralising and facilitating access to complex and dispersed information.

The use of tools such as Google Colab and Markmap has enabled better organisation and visualisation of data, increasing efficiency in knowledge management. This approach not only improves the understanding and use of critical mineral information, but also prepares users to apply this knowledge in real-world contexts.

The practical experience gained in this exercise is directly applicable to other projects that require customisation of language models for specific use cases.

8. Do you want to do the exercise?

If you want to replicate this exercise, access this this repository where you will find more information (the prompts used, the code generated by MateriaGuru, etc.)

Also, remember that you have at your disposal more exercises in the section "Step-by-step visualisations".

Content elaborated by Juan Benavente, industrial engineer and expert in technologies linked to the data economy. The contents and points of view reflected in this publication are the sole responsibility of the author.

In the vast technological landscape, few tools have made as deep a mark as Google Maps. Since its inception, this application has become the standard for finding and navigating points of interest on maps. But what happens when we look for options beyond the ubiquitous map application? In this post we review possible alternatives to the well-known Google application. 

Introduction 

At the beginning of 2005, Google's official blog published a brief press release in which they presented their latest creation: Google Maps. To get an idea of what 2005 was like, technologically speaking, it is enough to look at the most groundbreaking mobile terminals that year: 

Imagen credits: Cinco móviles que marcaron el año 2005 

Some of us still remember what the experience (or lack of experience) of running apps on these terminals was like. Well, in that year the first version of Google Maps was launched, allowing us to search for restaurants, hotels and other elements near our location, as well as to find out the best route to go from point A to point B on a digital version of a map of our city. In addition, that same year, Google Earth was also launched, which represented a real technological milestone by providing access to satellite images for almost all citizens of the world.   

Since then, Google's digital mapping and navigation ecosystem, with its intuitive interface and innovative augmented reality features, has been a beacon guiding millions of users on their daily journeys.

But what if we are looking for something different? What alternatives are there for those who want to explore new horizons? Join us on this journey as we venture into the fascinating world of your competitors. From more specialized options to those that prioritize privacy, we will discover together the various routes we can take in the vast landscape of digital navigation.

Alternatives to Google Maps  

Almost certainly some of you readers have seen or used some of the open source alternatives to Google Maps, although you may not know it. Just to mention some of the most popular alternatives:  

1. OpenStreetMap (OSM): OpenStreetMap is a collaborative project that creates a community-editable map of the world. It offers free and open geospatial data that can be used for a variety of applications, from navigation to urban analysis.  

2. uMap: uMap is an online tool that allows users to create custom maps with OpenStreetMap layers. It is easy to use and offers customization options, making it a popular choice for quick creation of interactive maps. 

3. GraphHopper: GraphHopper is an open source routing solution that uses OpenStreetMap data. It stands out for its ability to calculate efficient routes for vehicles, bicycles and pedestrians, and can be used as part of custom applications.  

4. Leaflet: Leaflet is an open source JavaScript library for interactive maps compatible with mobile devices. It is probably the most widespread library because of its low KB weight and because it includes all the mapping functions that most developers might need.  

5. Overture Maps: While the previous four solutions are already widely established in the market, Overture Maps is a new player. It is a collaborative project to create interoperable open maps. 

Of all of them, we are going to focus on OpenStreetMap (OSM) and Overture Maps.

Open Street Maps: an open and collaborative tool  

Of the aforementioned solutions, probably the most widespread and well-known is Open Street Maps.   

OpenStreetMap (OSM) stands out as one of the best open source alternatives to Google Maps for several reasons:   

• First, the fundamental characteristic of OpenStreetMap lies in its collaborative and open nature, where a global community contributes to the creation and constant updating of geospatial data. 

• In addition, OpenStreetMap provides free and accessible data that can be used flexibly in a wide range of applications and projects. To quote verbatim from their website: OpenStreetMap is open data: you are free to use it for any purpose as long as you credit OpenStreetMap and its contributors. If you modify or build upon the data in certain ways, you may distribute the result only under the same license. See the Copyright and License page for more details.  

• The ability to customize maps and the flexibility of OpenStreetMap integration are also outstanding features. Developers can easily tailor maps to the specific needs of their applications by leveraging the OpenStreetMap API. This is the key to the development of an ecosystem of applications around OSM such as uMap, Leaflet or GraphHopper, among many others. 

Overture Maps. A unique competitor  

Perhaps, one of the most promising projects to have recently appeared on the global technology scene is Overture Maps. As indicated (last July of this year) by its foundation (OMF Overture Maps Foundation), it has released its first open dataset, marking a significant milestone in the collaborative effort to create interoperable open map products. The first Overture release includes four unique data layers:   

• Places of Interest (POIs)  

• Buildings  

• Transportation Network  

• Administrative Boundaries 

Example coverage of public places worldwide identified in the initial project dataset. The first version of the overture maps dataset contains, among others, 59 million records of points of interest, 780 million buildings, transport networks and national and regional administrative boundaries worldwide. 

These layers, which merge various open map data sources, have been validated and contrasted through quality checks and are released under the Overture Maps data schema, made public in June 2023. Specifically, the Places of Interest layer includes data on more than 59 million places worldwide. This dataset is presented as a fundamental building block for navigation, local search and for various location-based applications. The other three layers include detailed building information (with more than 780 million building footprints worldwide), a global transportation network derived from the OpenStreetMap project, and worldwide administrative boundaries with regional names translated into more than 40 languages. 

Perhaps one of the most significant pieces of information in this announcement is the number of collaborators that have come together to realize this project. The Overture collaboration, founded in December 2022 by Amazon Web Services (AWS), Meta, Microsoft and TomTom, now boasts more than a dozen geospatial and technology companies, including new members such as Esri, Cyient, InfraMappa, Nomoko, Precisely, PTV Group, SafeGraph, Sanborn and Sparkgeo. The central premise of this collaboration is the need to share map data as a common asset to support future applications.  

As a good open source project, the Overture Foundation has made available to the development community a Github repository where they can contribute to the project.

In short, digital maps, their corresponding geospatial data layers, navigation and photo-geolocation capabilities are vital and strategic assets for social and technological organizations around the world. Now, with the 20th anniversary of the birth of Google Maps just around the corner, there are good open source alternatives and the big players in the international technology landscape are coming together to generate even more valuable spatial assets. Who will win this new race? We don't know, but we will keep a close eye on the current news on this topic.

The Canary Islands Statistics Institute (ISTAC) has taken a significant step forward in the volume of geographic data thanks to the publication of a total of 4,002 new datasets (3,859 thematic maps and 143 statistical cubes) in datos.gob.es, following its federation in Canarias Datos Abiertos.     

 This type of initiative is aligned with the European Union's Data Strategy, which establishes the guidelines to achieve a single data market that benefits companies, researchers and public administrations. The automation of publication processes through common standards is key to ensure interoperability and adequate access to open data sets of public administrations.

The generation of these datasets is the culmination of an automation work that has allowed the expansion of the number of published cubes, as now combinations of granularity and year since 2004 are presented.  In early October, the ISTAC added to its catalog more than 500 semantic assets and more than 2100 statistical cubes, as we told in this post on datos.gob.es. 

 In addition, the sets published to date have undergone a renewal process to become the aforementioned 143 statistical cubes.  The increase of these datasets not only improves the directory of datos.canarias.es and datos.gob.es in quantitative terms, but also broadens the uses it offers thanks to the type of information added.    

The indicators of these cubes are represented on the cartography through choropleth maps and in multiple formats. This automation will, in turn, not only allow other datasets to be published more easily in the future, but also more frequently.  

Another of the advances of this work is that the services are generated on the fly from the Geoserver map server, and not through an upload to CKAN, as was done until now, which reduces their storage and speeds up their updating. 

How to bring demographic indicators closer to the population   

Demographic indicators are dense data cubes that offer a large amount of detailed geographic information, including total population, disaggregated by sex, residence, age and other indices up to a total of 27 different variables. 

As so much information is contained in each cube, it can be difficult to represent specific indicators on the cartography, especially if the user is not used to working with certain GIS (Geographical Information System) software. 

To bring this content to all types of users, the ISTAC has generated 3,859 new maps, representing on a choropleth map each of the indicators contained in the 143 statistical cubes. The publication of these new cartographic data is thus presented as a more efficient and simplified way of obtaining the information already represented, allowing users to easily access the specific data they need.

We could compare this transformation to flowers.  Previously, only whole bouquets were published, with 27 flowers per bouquet, which had to be managed and handled to represent the flowers that were of interest. Now, in addition to continuing to publish the bouquets, new processes have been generated to be able to publish each flower separately, automating the generation of each of these sets, which will also be updated more frequently.

This new option facilitates the use of these choropleth maps (like the one shown in the image) by people without technical GIS knowledge, since they are presented in easily downloadable formats as images (.jpg and .png) for professional, educational or personal use.

      Mapa de población de 65 o más años (% sobre total) por municipios. Año 2022

For more advanced users, ISTAC has also expanded the range of formats in which the original indicator cubes are served. The "bouquets", which previously only showed data in CSV format, now have a wide variety of distributions: KML, GML, GeoPackage, GeoJSON, WFS, WMS. Taking advantage of the benefits provided by the use of styles in the WMS format, all the styles associated with the indicators have been generated, so that, using them, it is possible to represent the same map that is downloaded in image format. These styles are calculated for each indicator-granularity-year combination, according to the method of calculating quantiles for five intervals. 

This new approach with both simple and complex geographic data enriches the catalog and allows users without specific knowledge to access and reuse them. In addition, it should be noted that this opens the door to other massive publications of data based on other statistical operations.

In short, this is an important step in the process of opening up data.  A process that improves the use and sharing of data, both for the user on the ground and for professionals in the sector. Given the growing need to share, process and compare data, it is essential to implement processes that facilitate interoperability and appropriate access to open data. In this sense, the Canary Islands Institute of Statistics is concentrating its efforts to ensure that its open data sets are accessible and in the appropriate formats for sharing. All this in order to obtain value from them.

Data is a valuable source of knowledge for society. Public commitment to achieving data openness, public-private collaboration on data, and the development of applications with open data are actions that are part of the data economy, which seeks the innovative, ethical, and practical use of data to improve social and economic development.

It is as important to achieve public access and use of data as it is to properly convey that valuable information. To choose the best chart for each type of data, it is necessary to identify the type of variables and the relationship between them.

When comparing data, we must ensure that the variables are of the same nature, i.e., quantitative or qualitative variables, in the same unit of measurement, and that their content is comparable.

We present below different visualizations, their usage rules, and the most appropriate situations to use each type. We address a series of examples, from the simplest ones like bar charts to less well-known charts like heat maps or stacked comparisons.

Bar charts

A visualization that represents data using two axes: one that collects qualitative or time data and another that shows quantitative values. It is also used to analyze trends because one of the axes can show temporal data. If the axes are flipped, a column chart is obtained.

Best practices:

• Display the axis value labels and reserve labels as tooltips for secondary data.
• Use it to represent less than 10 value points. When visualizing more value points, a line chart may be more appropriate.
• Clearly differentiate real data from estimates.
• Combine with a line chart to show trends or averages.
• Place the one with longer descriptions on the vertical axis, when no variable is temporal.

Source: El Orden Mundial https://elordenmundial.com/mapas-y-graficos/comercio-fertilizantes-mundo/

Clustered bar charts

A type of bar chart in which each data category is further divided into two or more subcategories. Therefore, the comparative scenario encompasses more factors.

Best practices

• Limit the number of categories to avoid showing too much information on the chart.
• Introduce a maximum of three or four subcategories within each category. In case more groupings need to be shown, the use of stacked bars or a set of charts can be considered.
• Choose contrasting colors to differentiate the bars of each subcategory.

Source: RTVE https://www.rtve.es/noticias/20230126/pobreza-energetica-espana/2417050.shtml

Cumulative comparison charts

These charts display the composition of a category in a cumulative manner. In addition to providing a comparison between variables, these charts can show the segmentation of each category. They can be either stacked bar charts or cumulative area charts.

Best practices

• Avoid using stacked bar charts when comparing segments of each category to each other. In that case, it is better to use multiple charts.
• Limit the number of subcategories in stacked bar charts or segments in area charts.
• Apply contrast in colors between categories and adhere to accessibility principles.

Source: Newtral https://www.newtral.es/medallas-espana-eurobasket/20220917/

Population pyramid

A combination of two horizontal bar charts that share a vertical axis representing the initial value and display two values that grow symmetrically on either side.

Best practices

• Define a common ordering criterion such as age.
• Represent the data in absolute numbers or percentages to take into account that the sum of the two values being compared represents the total.

Source: El Español https://www.elespanol.com/quincemil/articulos/actualidad/asi-es-la-alarmante-piramide-de-poblacion-de-galicia-en-2021

Radar chart

Circular visualization formed by polar axes that are used to represent measurements with categories that are part of the same theme. From each category, radial axes converge at the central point of the chart.

Good practices:

• Keep numerical data within the same range of values to avoid distorting a chart.
• Limit the number of categories in data series. An appropriate number could be between four and seven categories.
• Group categories that are related or share a common hierarchy in one sector of the radar chart.

Source: Guía de visualización de datos para Entidades Locales https://redtransparenciayparticipacion.es/download/guia-de-visualizacion-de-datos-para-entidades-locales/

Heatmap

A graphical representation in table format that allows for the evaluation of two different dimensions differentiated by degrees of color intensity or traffic light codes.

Good practices:

• Indicate the value in each cell because color is only an indicative attribute. In interactive graphics, values can be identified with a pop-up label.
• Include a scheme or legend in the graphic to explain the meaning of the color scale.
• Use accessible colors for everyone and with recognizable semantics such as gradients, hot-cold, or traffic light colors.
• Limit or reduce the represented information as much as posible.

Source: eldiario.es https://www.eldiario.es/sociedad/clave-saturacion-primaria-ratios-mitad-medicos-asignados-1-500-pacientes_1_9879407.html

Bubble chart

A variation of the scatter plot that, in addition, represents an additional dimension through the size of the bubble. In this type of chart, it is possible to assign different colors to associate groups or separate categories. Besides being used to compare variables, the bubble chart is also useful for analyzing frequency distributions. This type of visualization is commonly found in infographics when it is not as important to know the exact data as it is to highlight the differences in the intensity of values.

Good practices:

• Avoid overlapping bubbles so that the information is clear.
• Display value labels whenever possible and the number of bubbles allows for it.

Source: Civio https://civio.es/el-boe-nuestro-de-cada-dia/2022/07/07/decretos-ley-desde-1996/

Word cloud

A visual graphic that displays words in varying sizes based on their frequency in a dataset. To develop this type of visualization, natural language processing (NLP) is used, which is a field of artificial intelligence that uses machine learning to interpret text and data.

Good practices:

• It is recommended to use this resource in infographics where showing the exact figure is not relevant but a visual approximation is.
• Try to make the length of the words similar to avoid affecting perception.
• Make it easier to read by showing the words horizontally.
• Present the words in a single color to maintain a neutral representation.

This graphic visualization, which we published in a step-by-step article, is a word cloud of several texts from datos.gob.es.

So far, we have explained the most common types of comparison charts, highlighting examples in media and reference sources. However, we can find more visualization models for comparing data in the Data Visualization Guide for Local Entities, which has served as a reference for creating this post and others that we will publish soon. This article is part of a series of posts on how to create different types of visualizations based on the relationship of the data and the objective of each exercise.

As the popular mantra goes, "a picture is worth a thousand words," which could be adapted to say that "a chart is worth a thousand numbers." Data visualization serves to make information understandable that, a priori, could be complex.

Talking about GPT-3 these days is not the most original topic in the world, we know it. The entire technology community is publishing examples, holding events and predicting the end of the world of language and content generation as we know it today. In this post, we ask ChatGPT to help us in programming an example of data visualisation with R from an open dataset available at datos.gob.es.

Introduction

Our previous post talked about Dall-e and GPT-3's ability to generate synthetic images from a description of what we want to generate in natural language. In this new post, we have done a completely practical exercise in which we ask artificial intelligence to help us make a simple program in R that loads an open dataset and generates some graphical representations.

We have chosen an open dataset from the platform datos.gob.es. Specifically, a simple dataset of usage data from madrid.es portals. The description of the repository explains that it includes information related to users, sessions and number of page views of the following portals of the Madrid City Council: Municipal Web Portal, Sede Electrónica, Transparency Portal, Open Data Portal, Libraries and Decide Madrid.

The file can be downloaded in .csv or .xslx format and if we preview it, it looks as follows:

OK, let's start co-programming with ChatGPT!

First we access the website and log in with our username and password. You need to be registered on the openai.com website to be able to access GPT-3 capabilities, including ChatGPT.

We start our conversation:

During this exercise we have tried to have a conversation in the same way as we would have with a programming partner. So the first step we do is to say ‘hello’ and mention the problem we have. When we ask the AI to help us create a small program in R that graphically represents some data, it gives us some examples and helps us with the explanation of the program:

Since we have no data, we cannot do anything practical at the moment, so we ask it to help us generate some synthetic data.

As we say, we behave with the AI as we would with a person (it looks good).

Once the AI seems to easily answer our questions, we go to the next step, we are going to give it the data. And here the magic begins... We have opened the data file that we have downloaded from datos.gob.es and we have copied and pasted a sample.

With the data copied and pasted as we have given it to it, the AI writes the code in R to load it manually into a dataframe called \"data\". It then gives us the code for ggplot2 (the most popular graphics library in R) to plot the data along with an explanation of how the code works.

Great! This is a spectacular result with a totally natural language and not at all adapted to talk to a machine. Let's see what happens next:

But it turns out that when we copy and paste the code into an RStudio environment it is no running.

So, we tell to it what's going on and ask it to help us to solve it.

We tried again and, in this case, it works!

However, the result is a bit clumsy. So, we tell it.

From here (and after several attempts to copy and paste more and more rows of data) the AI changes the approach slightly and provides me with instructions and code to load my own data file from my computer instead of manually entering the data into the code.

We take its opinion into account and copy a couple of years of data into a text file on our computer. Watch what happens next:

We try again:

As you can see, it works, but the result is not quite right.

And let's see what happens.

Finally, it looks like it has understood us! That is, we have a bar chart with the visits to the website per month, for the years 2017 (blue) and 2018 (red). However, I am not convinced by the format of the axis title and the numbering of the axis itself.

Let's look at the result now.

It looks much better, doesn't it? But what if we give it one more twist?

However, it forgot to tell us that we must install the plotly package or library in R. So, we remind it.

Let's have a look at the result:

As you can see, we have now the interactive chart controls, so that we can select a particular year from the legend, zoom in and out, and so on.

Conclusion

You may be one of those sceptics, conservatives or cautious people who think that the capabilities demonstrated by GPT-3 so far (ChatGPT, Dall-E2, etc) are still very infantile and impractical in real life. All considerations in this respect are legitimate and, many of them, probably well-founded.

However, some of us have spent a good part of our lives writing programs, looking for documentation and code examples that we could adapt or take inspiration from; debugging bugs, etc. For all of us (programmers, analysts, scientists, etc.) to be able to experience this level of interlocution with an artificial intelligence in beta mode, made freely available to the public and being able to demonstrate this capacity for assistance in co-programming, is undoubtedly a qualitative and quantitative leap in the discipline of programming.

Content prepared by Alejandro Alija, Digital Transformation expert.

The contents and points of view reflected in this publication are the sole responsibility of the author.

A statistical graph is a visual representation designed to contain a series of data whose objective is to highlight a specific part of the reality. However, organising a set of data in an informative way is not an easy task, especially, if we want to capture the viewer’s attention and to present the information in an accurate format.

In order to facilitate comparisons between data it is required a minimum of statistical knowledge to highlight trends, to avoid misleading visualisation and to illustrate the message to be conveyed. Therefore, depending on the type of interrelation that exists between the data that we are trying to illustrate, we must choose one type of visualisation or another. In other words, representing a numerical classification is not the same as representing the degree of correlation between the two variables.

In order to precisely choose the most appropriate graphs according to the information, we have selected down the most recommended graphs for each type of association between numerical variablesv. During the process of preparing this content, we have taken as a reference the Data Visualisation Guide for local entities recently published by the FEMP's RED de Entidades Locales por la Transparencia y Participación Ciudadana, as well as this infography prepared by the Financial Times.

Deviation

It is used to highlight numerical variations from a fixed reference point. Usually, the reference point is zero, but it can also be a target or a long-term average. In addition, this type of graph is useful to show sentiments (positive, neutral or negative). The most common charts are:

• Diverging bar: A simple standard bar chart that can handle both negative and positive magnitude values.
• Column chart: Divides a single value into 2 contrasting components (e.g. male/female).

Correlation

Useful to show the relationship between two or more variables. Note that unless you tell them otherwise, many readers will assume that the relationships you show them are causal. Here are some of the graphs.

• Scatter plot: The standard way of showing the relationship between two continuous variables, each of which has its own axis.
• Timeline: A good way to show the relationship between a quantity (columns) and a ratio (line).

Sorting

Sorting numerical variables is necessary when the position of an item in an ordered list is more important than its absolute or relative value. The following graphs can be used to highlight points of interest.

• Bar chart: These types of visualisations allow ranges of values to be displayed in a simple way when they are sorted.
• Dot-strip chart: The values are arranged in a strip. This layout saves space for designing ranges in multiple categories.

Distribution

This type of graph seeks to highlight a series of values within a data set and represent how often they occur. That is, they are used to show how variables are distributed over time, which helps to identify outliers and trends.

The shape itself of a distribution can be an interesting way to highlight non-uniformity or equality in the data. The most recommended visualisations to represent, for example, are age or gender distribution are as follows:

• Histogram: This is the most common way of showing a statistical distribution. To develop it, it is recommended to keep a small space between the columns in order to highlight the "shape" of the data.
• Box plot: Effective for visualising multiple distributions by showing the median (centre) and range of the data.
• Population pyramid: Known for showing the distribution of the population by sex. In fact, it is a combination of two horizontal bar charts sharing the vertical axis.

Changes over time

Through this combination of numerical variables it is possible to emphasise changing trends. These can be short movements or extended series spanning decades or centuries. Choosing the right time period to represent is key to providing context for the reader.

• Line graph: This is the standard way to show a changing time series. If the data is very irregular it can be useful to use markers to help represent data points.
• Calendar heat map: Used to show temporal patterns (daily, weekly, monthly). It is necessary to be very precise with the amount of data.

Magnitude

It is useful for visualising size comparisons. These can be relative (simply being able to see bigger/larger) or absolute (requires seeing more specific differences). They usually show variables that can be counted (e.g. barrels, dollars or people), rather than a calculated rate or percentage.

• Column chart: One of the most common ways to compare the size of things. The axis should always start at 0.
• Marimekko chart: Ideal for showing the size and proportion of data at the same time, and as long as the data is not too complex.

Part of a whole

These types of numerical combinations are useful to show how an entity itself can be broken down into its constituent elements. For example, it is common to use part of a whole to represent the allocation of budgets or election results.

• Pie chart: One of the most common charts to show partial or complete data. Keep in mind that it is not easy to accurately compare the size of different segments.
• Stacked Venn: Limited to schematic representations to show interrelationships or coincidences.

Spatial

This type of graph is used when precise locations or geographic patterns in the data are more important to the reader than anything else. Some of the most commonly used are:

• Choropleth map: This is the standard approach to placing data on a map.
• Flow map: This is used to show movement of any kind within a single map. For example, it can be used to represent migratory movements.

By knowing the different statistical representation options, it helps to create more accurate data visualisations, which in turn to allow a more clearly conceived reality. Thus, in a context where visual information is becoming increasingly important, it is essential to develop the necessary tools so that the information contained in the data reaches the public and contributes to improving society.

The FEMP's Network of Local Entities for Transparency and Citizen Participation has just presented a guide focused on data visualisation. The document, which takes as a reference the Guide to data visualisation developed by the City Council of L'Hospitalet, has been prepared based on the search for good practices promoted by public and private organisations.

The guide includes recommendations and basic criteria to represent data graphically, facilitating its comprehension. In principle, it is aimed at all the entities that are members of the FEMP's Network of Local Entities for Transparency and Citizen Participation. However, it is also useful for anyone wishing to acquire a general knowledge of data visualisation.

Specifically, the guide has been developed with three objectives in mind:

• To provide principles and good practices in the field of data visualisation.
• To provide a model for the visualisation and communication of local authority data by standardising the use of different visual resources.
• Promote the principles of quality, simplicity, inclusiveness and ethics in data communication.

What does the guide include?

After a brief introduction, the guide begins with a series of basic concepts and general principles to be followed in data visualisation, such as the principle of simplification, the use of space or accessibility and exclusive design. Through graphic examples, the reader learns what to do and what not to do if we want our visualisation to be easily understood. 

The guide then focuses on the different stages of designing a data visualisation through a sequential methodological process, as shown in the following diagram:

As the image shows, before developing the visualisation, it is essential to take the time to establish the objectives we want to achieve and the audience we are targeting, in order to tailor the message and select the most appropriate visualisation based on what we want to represent.

When representing data, users have at their disposal a wide variety of visualisation objects with different functions and performance. Not all objects are suitable for all cases and it will be necessary to determine the most appropriate one for each specific situation. In this sense, the guide offers several recommendations and guidelines so that the reader is able to choose the right element based on his or her objectives and audience, as well as the data he or she wants to display.

The following chapters focus on the various elements available (infographics, dashboards, indicators, tables, maps, etc.) showing the different subcategories that exist and the good practices to follow in their elaboration, showing numerous examples that facilitate their understanding. Recommendations on the use of the text are also provided.

The guide ends with a selection of resources for further knowledge and data visualisation tools to be considered by anyone who wants to start developing their own visualisations.

You can download the complete guide below, in the "Documentation" section.