Showcases on Data Toolbox

Coloring Administrative District Maps with Data

Mon, 01 Jan 0001 00:00:00 +0000

Coloring administrative district maps — commonly known as blank maps — with data is called a “choropleth map.” Although sometimes referred to as a “classified area map,” that term actually describes a data processing technique, not a visualization type.

Using our “Japan Choropleth Map” and “Prefecture Choropleth Map” tools, you can easily create high-quality choropleth maps.

Here we introduce how to create one using our tools.

dataviz.jp

Fukushima Prefecture: Population Change Rate 2011–2024

choropleth-prefectures.dataviz.jp

Paid subscribers can open the link above as an editable project file to learn the process and examine the data structure.

Visualizing with Sample Data

Here is what it looks like when you visualize pre-loaded population data on a Fukushima Prefecture map. By turning on the “Divide by area (/km²)” checkbox, the data becomes population density.

Incidentally, using a technique called a cartogram — which distorts area to reflect data (also available as a tool) — produces a warped map like this:

Visualizing with Your Own Data

You can of course prepare and visualize your own data.

In that case, download the sample data and keep only the place names to see what geographic names you need to prepare data for.

Population Comparison Data Between 2011 and 2024

Download the “Fukushima Prefecture Current Population Survey Annual Report: Reference Materials” published by Fukushima Prefecture. This dataset compares population between 2011 and 2024.

Keep only the sheet and table you need.

Once you have this, load it into OpenRefine for cleansing.

Remove unnecessary spaces and commas. The minus sign is represented as “△”, so replace it with the minus symbol “-”.

Also load the file with only municipality names from the sample data into OpenRefine as a separate project.

Use the “Add column based on this column” operation to perform a VLOOKUP-like join — merging two tables horizontally.

Join using an OpenRefine-specific script:

cell.cross('PROJECT_NAME','KEY_COLUMN_NAME').cells['COLUMN_NAME_TO_GET'].value[0]

PROJECT_NAME — the project name of your original data
KEY_COLUMN_NAME — the name of the column to use as the key
COLUMN_NAME_TO_GET — the name of the column you want to join

You can see that the join was successful for all municipalities except Futaba, Namie, Okuma, and Tomioka, where no data exists.

Municipalities with higher population change rates are displayed in darker colors.

Exploring Two Variables at Once with a Mekko Chart

Mon, 01 Jan 0001 00:00:00 +0000

Every year, Dentsu publishes “Advertising Expenditures in Japan.” We visualized the advertising expenditure by industry across four traditional media (excluding the internet).

Normally, using bar charts or pie charts would require creating as many charts as there are industries (21) or media types (4). With a Mekko chart, however, you can represent both dimensions as vertical and horizontal proportions within a single chart.

Here we introduce how to create one using the tools provided by this service.

dataviz.jp

Advertising Expenditure by Industry and Mass Media (2025)

rawgraphs.dataviz.jp

Paid users can open the link above as an editable project file to learn how it was made and examine the data structure.

The data is sourced from Dentsu’s PDF report.

2025 Advertising Expenditures in Japan - Dentsu

Extracting Table Data from a PDF

The published materials are available as a PDF. As-is, this data cannot be processed by a computer. Manual data entry would be tedious.

In cases like this, Tabula makes it easy to extract table data from PDFs.

dataviz.jp

Tabula PDF

Extract table data from PDFs

tabula-pdf.dataviz.jp

Cleansing the Data

To transform the data into a tidy format that computers can process, OpenRefine is very useful. It allows you to cleanse data with visual confirmation.

We removed unnecessary columns, as well as unnecessary spaces and commas.

The data is still in a cross-tabulation (matrix) format, so we need to perform the reverse of a pivot table in Excel. OpenRefine lets you do this visually as well.

Data cleansing is now complete. Choose your favorite visualization tool. In this case, we’ll use RawGraphs, which makes it easy to create Mekko charts.

Visualizing Data with RawGraphs

RawGraphs is a single-page tool where each step appears as you scroll down.

Completed

Export as PNG to include in slides or publish on the web.

Export as SVG to edit in design tools such as PowerPoint, Adobe Illustrator, or Figma. Some axis labels were too close together, so we adjusted them.

Editing in PowerPoint

Editing in Adobe Illustrator

Finding the Ideal Yogurt with Parallel Coordinates

Mon, 01 Jan 0001 00:00:00 +0000

Let’s find the “ideal yogurt” — high in protein, low in sugar.

With a Parallel Coordinates chart, you can easily filter data by applying multiple conditions simultaneously.

Here we show how to create this using our tools.

dataviz.jp

Finding the Ideal Yogurt with Parallel Coordinates

parallel-coordinates.dataviz.jp

Paid subscribers can open the link above as an editable project file to learn the process and examine the data structure.

Collecting the Data

First, we gathered nutritional information for yogurts sold at convenience stores, collecting data from each manufacturer’s website.

Visualizing with Parallel Coordinates

After loading nutritional data for 70 yogurt products and switching to Min-Max scale, metrics with different units are normalized to the same 0–1 axis.

Filtering Across Multiple Axes

By brushing the upper range on the protein axis and the lower range on the carbohydrate axis, only products meeting the “high protein, low sugar” criteria remain as blue lines.

The standout result is PARTHENO Plain, a rich Greek yogurt — 10.9g of protein with just 4.6g of carbohydrates.

Reordering Axes

Try dragging axes to rearrange them, and hover over table rows to highlight the corresponding lines. Experience how the “feel” of the data changes with different arrangements.

Visualizing Japan's Temperature Changes with a Spiral Chart

Mon, 01 Jan 0001 00:00:00 +0000

We visualized the annual mean temperature anomaly in Japan from 1898 to the present using a spiral chart.

With RawGraphs’ spiral chart, you can arrange long-term time series data in a spiral layout, making it easy to grasp trends over time at a glance.

Here we introduce how to create this using our service’s tools.

dataviz.jp

Visualizing Japan's Temperature Changes with a Spiral Chart

rawgraphs.dataviz.jp

Data Source

This visualization uses the “Annual Mean Temperature Anomaly in Japan” data published by the Japan Meteorological Agency (JMA).

Annual Mean Temperature Anomaly in Japan — Japan Meteorological Agency

This dataset records the temperature anomaly (°C) for each year from 1898 to 2025, calculated against the 30-year average from 1991 to 2020 as the baseline. In recent years, positive anomalies exceeding +1°C have continued, clearly showing a long-term warming trend.

Visualization with RawGraphs

Load the CSV data into RawGraphs and select “Spiral Plot” from the chart types.

By assigning the year to the time axis and the temperature anomaly to the value, a spiral chart is drawn with the timeline progressing from the center outward. The closer to the outer edge, the more recent the data, and the warming trend becomes visually apparent through changes in color and position.

Advantages of Spiral Charts

While bar charts or line charts can represent the same data, they become too wide for long-term data spanning over 100 years. Spiral charts can fit long time series into a compact area, making it easier to get an overview of the overall trend.

By comparing the inner part of the spiral (past) with the outer part (present), you can intuitively grasp how the temperature anomaly has been growing larger year by year.