This dataset contains 500 entries of housing price data from various countries, regions, and cities worldwide, making it ideal for machine learning models and real estate market analysis. The dataset covers diverse geographic locations, including:

North America: USA, Canada, Mexico
Europe: Germany, France, UK, Italy, Spain
Asia: Japan, China, India, South Korea
Other Regions: Australia, Brazil, South Africa

Columns Included:

Country: The country where the house is located (e.g., USA, Japan, India).
State/Region: The state or region within the country (e.g., California, Bavaria).
City: The city where the property is located (e.g., Los Angeles, Tokyo).
Square Footage (SqFt): The size of the house in square feet (ranging from 500 to 5000 sq ft).
Bedrooms: The number of bedrooms in the house (ranging from 1 to 6).
Population Density: The population density of the area (people per sq km).
Price of House: The price of the house (in local currency, converted to USD where applicable).

This dataset can be used for:

Machine Learning Models: Training and evaluating models for house price prediction.
Market Analysis: Analyzing housing trends across different regions and countries.
Visualization: Creating insightful visualizations to understand price distributions and regional variations.

This dataset provides a balanced mix of geographic diversity and housing features for robust predictive modeling and analysis.

Detailed methods can be found in the publication, and highlights are provided below. The following original data sources were aggregated/disaggregated to a common hexagonal grid (cell size 290 km2, mean internode spacing 18.3 km): Gridded Livestock of the World (GLW 2), doi:10.1371/journal.pone.0096084, reporting year 2006, resolution 3 arcminutes (~5 km2 at equator); Gridded Population of the World (GPWv4), doi:10.7927/H4HX19NJ, reporting year 2010, resolution 30 arcseconds (~1 km at equator); GlobCover 2009, doi:10.1594/PANGAEA.787668, reporting year 2009, resolution 300m; FAOSTAT Fertilizers by Nutrient dataset (downloaded on 26 Feb 2018), http://www.fao.org/faostat/en/#data/RFN/metadata, reporting years 2002-2014, resolution national. ---Subnational methods and calculations Livestock densities, human population density, and cropland extent were summarized for each grid cell in a global hexagonal grid. This grid had consistent grid cell areas across latitudes, and was generated using the dggrid package (Barnes, 2016; Sahr, 2011) in the platform R (R Core Team, 2016). In the finer hexagonal grid, each grid cell had a mean area of 290 km2 and a mean internode spacing of 18.3 km. In the coarser grid, each grid cell had a mean side length of 95 km (mean hexagon area of 23,300 km2, mean internode spacing of 165 km), which was large enough to encompass megacities such as London and Paris along with peri-urban areas, but small enough to maintain subnational resolution in relatively small nations. For a minority of hexagonal grid cells, slight deviations in the dimensions were mathematically necessary to avoid overlapping cells and gaps over the world's surface (Barnes, 2016). Total manure P production in each grid cell was calculated by summing the contributions from each animal type, using animal-specific and nation-specific P excretion factors from Bouwman et al. (2017). For cattle we used 16.6 kg P per head yr-1 in Canada, USA, and Japan, 13.1 kg P per head yr-1 in the other OECD countries, and 8.75 kg P per head yr-1 in the remaining countries (Bouwman et al. 2017). For other animals we used 1.8 kg P per head yr-1 for pigs, 0.1 kg P per head yr-1 for chickens, 1.5 kg P per head yr-1 for sheep and goats for all countries (Bouwman et al. 2017). Cells with zero cropland extent were excluded from the analysis (and thus also gridcelldata.csv). --National methods and calculations We used nation-level P fertilizer data from FAOSTAT including import, export, agricultural use, and production for the most recent available years (2002-2014). FAOSTAT data were downloaded on 26 Feb 2018. Fertilizer data are reported annually, and we took the nation-specific means for each budgetary term over two different five year intervals (2010-2014, 2002-2006); these years deliberately exclude the global food crisis of 2007/2008 when the global phosphate rock price spiked by 400% (Chowdhury et al., 2017). A small number of countries had data gap years, requiring that the mean be calculated over fewer years. Import ratios, an indicator of fertilizer P import dependency, were calculated as net import : consumption, where net import = import - export. Recent fertilizer P consumption trends were summarized by calculating a consumption ratio of the 2010s to 2000s (2010-2014:2002-2006). Calculations involving P import ratios and consumption trends were conducted directly on FAO data, prior to disaggregation within the global grid. In cases where grid cells overlapped multiple countries, the nation representing the largest share of the grid cell was assigned to the whole cell using administrative data from Natural Earth. A minority of nations lacked P import or P consumption data and were excluded from P import ratio calculations. Nations that lacked P export data were assumed to have zero gross P export in these calculations. Attributes of the two compiled subnational datasets: "gridcelldata_fine.csv" and "gridcelldata_coarse.csv" (each row represents one hexagonal grid cell) nation: Name of the nation that possessed the largest share of the grid cell. lat: Decimal latitude of the grid cell centroid. lon: Decimal longitude of the grid cell centroid. crop_pct: Mean percent of land as cropland (i.e., cropland extent) within the grid cell. For coastal grid cells, only the land portion of the cell was used in this calculation. popd_indperkm2: Mean population density of the grid cell. manurep_kgperkm2: Calculated manure P production of the grid cell. This is the sum across multiple animal types using animal-specific, nation-specific P excretion factors from Bouwman et al. 2017. cattle_indperkm2: Mean cattle density of the grid cell. pigs_indperkm2: Mean pig density of the grid cell. chickens_indperkm2: Mean chicken density of the grid cell. sheep_indperkm2: Mean sheep density of the grid cell. goats_indperkm2: Mean goat density of the grid cell. pfertnatcons10s_metrictons: Mean nation-level P fertilizer consumption (P2O5 total nutrients) for years 2010-2014, for the nation that possessed the largest share of the grid cell. Calculated from FAOSTAT. pfertnatimp_metrictons: Mean nation-level P fertilizer import (P2O5 total nutrients) for years 2010-2014, for the nation that possessed the largest share of the grid cell. Calculated from FAOSTAT.pfertnatout_metrictons: Mean nation-level P fertilizer export (P2O5 total nutrients) for years 2010-2014, for the nation that possessed the largest share of the grid cell. Calculated from FAOSTAT. pfertnatnetimpratio_unitless: Nation-level net fertilizer P import ratios ([import-export]/consumption) for years 2010-2014, for the nation that possessed the largest share of the grid cell. Calculated from FAOSTAT. pfertnatcons00s_metrictons: Mean nation-level P fertilizer consumption (P2O5 total nutrients) for years 2002-2006, for the nation that possessed the largest share of the grid cell. Calculated from FAOSTAT. pfertnatconstr_unitless: Nation-level P fertilizer consumption trend, for the nation that possessed the largest share of the grid cell. This is the ratio of 2010s:2000s (that is, mean of 2010-2014 divided by mean of 2002-2006). Calculated from FAOSTAT.