During Taylor Swift's European tour, we investigated why Spanish audiences found it difficult to remember her songs. The study identified two primary reasons: her unusually long verses compared to other pop artists like Ariana Grande, and her minimalist musical style, analyzed through Spotify's music metrics. We also pinpointed two key factors behind her success: the relatable themes in her lyrics that resonate with audiences, and her self-referential songwriting, which creates a unique "song universe" for fans. This combination of musical complexity and emotional connection explains her widespread popularity despite the challenges some listeners face in recalling her lyrics.

Data were manually extracted from Genius, but all datasets were analyzed with Python, particularly for text analysis. Libraries like re, pandas, and numpy were used to break down the words and verses. To analyze the musical metrics, Python and Spotify's API were employed. Several visualizations were created using the ggplot2 library in R, providing a clear representation of the song structures and patterns that contribute to Taylor Swift's style and success. Link to article

We examined Europe's population density using a detailed grid of one square kilometer, focusing on how population distribution has shifted over the last decade. Although this study was part of the broader Pulse program with European journalists, we centered on Spain. The article highlighted factors like average population age, dependency ratios, unemployment rates, and income levels to explain demographic trends. It offered insight into regional differences and urbanization patterns. The data allowed us to discuss how various socio-economic factors influence where and how people live in Spain, as well as broader implications for infrastructure, services, and policy planning across the country.

We worked with two population grids from Eurostat (2011 and 2021), enabling comparisons over a decade. These files were handled using Python and geopandas. Since the grids only contained population data per square kilometer, additional data from Eurostat on European municipalities were integrated through spatial joins, also using geopandas. The main map was created using Mapbox GL JS, while static maps were generated in QGIS and Mapbox. This allowed for a detailed spatial analysis of population changes and trends across Spain and Europe. Link to article

Last summer's drought in Spain led to severe water restrictions for farmers, with public administrations limiting water supply to critical levels. This article explored the administrative mismanagement of public water resources, exacerbating the drought's effects. Additionally, it discussed how excessive water usage by farmers during wetter periods contributes to resource depletion in times of scarcity. The combination of poor regulation and unsustainable agricultural practices has deepened the crisis. The article emphasized the need for improved water governance and sustainable resource management to mitigate future impacts of climate change, ensuring both agricultural viability and the long-term preservation of water supplies.

Data for this article came from the Citizen Drought Observatory and the ESYRCE survey by the Ministry of Agriculture. Flourish and Datawrapper were used to create the visuals and charts, making the information accessible and easy to understand. Link to article

After a casual conversation at work, we decided to see if we could predict the behavior of contestants on the reality show La Isla de las Tentaciones based on their language use in early episodes. We identified key repeated words and phrases and analyzed how these were connected to certain behaviors. The study found specific words that could potentially indicate cheating within couples. By mapping out the most common language patterns, we aimed to predict future actions on the show. This exploration offered insight into how language reflects behavior and relationships, providing a fun yet analytical take on reality TV dynamics.

We downloaded all episodes and converted the audio into subtitles. Python was used with the webvtt library to open the subtitles and nltk for natural language processing. This allowed us to identify frequently repeated terms and analyze sentence structures. Further manual processes assigned phrases to individual contestants. Most of the visualizations were designed in Figma, with a few created in R using the ggplot2 library. This approach enabled a comprehensive linguistic analysis of the reality show's dynamics. Link to article

Using data from MITMA and ECEPOV, we created a map of Spain’s regular commuting routes for work and study, analyzing the time spent traveling. The article highlighted how the COVID-19 pandemic had altered these patterns. It found that most Spaniards commute by car, often leading them to live on the outskirts of cities. In some cases, high-speed trains (AVE) played a significant role in long-distance commuting. The findings underscored how urban sprawl and a reliance on cars shaped commuting behaviors in Spain. The data also hinted at the broader effects of remote work trends and their impact on future mobility needs.

We worked with a CSV file containing commuting data and a geojson file with municipal centroids. These datasets were merged to create connections between municipalities, with Python used to extract lines representing commuting routes and calculate straight-line distances. The primary map was built using Mapbox GL JS, with filters based on travel distances, while additional charts were generated using Datawrapper, enhancing the visualization of commuting patterns across Spain. Link to article

The Ministry of Housing, in collaboration with a consulting firm, released a tool showing the heating demand for every building in Spain. We used this data to create our own map, incorporating the cost of renovations that would improve energy efficiency in older buildings. The article emphasized that Spain's housing stock is outdated, making it expensive to retrofit for better energy efficiency. The report called attention to the need for government support and awareness to tackle these challenges, especially as energy costs rise. It also discussed the environmental and economic benefits of upgrading Spain’s aging buildings to meet modern energy standards.

We scraped the property cadastre webpage using Python's beautiful soup to obtain building data, while heating demand data was extracted from the URBAN3R viewer, also through scraping. The datasets were combined using geopandas, but the large file size required optimization with Mapshaper and other command-line libraries. Additional charts were generated in Datawrapper using data from the 2011 Population and Housing Census, providing a comprehensive analysis of energy efficiency in Spain's building stock. Link to article

With data from AEMET, we analyzed how often air conditioning was used across Spain's provinces during the spring and summer of 2023. Over 122 days, Sevilla emerged as the hottest region, experiencing 82 days with temperatures exceeding 32°C. The study also included future temperature projections using a 20-kilometer grid, showing how heatwaves may intensify in the coming years. The article highlighted the growing need for cooling in homes and businesses due to rising temperatures, pointing to the broader issue of climate change. It also touched on the economic and environmental challenges posed by the increased reliance on air conditioning in Spain.

Temperature projections were retrieved via Aemet’s API using Python, while temperature data were queried directly from their database using SQL. Visuals were created in Datawrapper, QGIS (with further editing in Figma for maps), and R (with Figma edits for evolution charts). This allowed for an in-depth examination of air conditioning use and future temperature trends across Spanish provinces. Link to article