Analyzing with the Chart Builder¶
Read how to configure visualizations for a dataset beforehand.
With the Chart Builder tool, you can visualize multiple datasets on a single customized chart. You can save your chart and easily share it, embed it, or use it as a widget.
It is usually made available by the domain administrator in the top menu link under the codename Charts. The Map Builder can also be accessed using its URL:
You can also build charts with Analyze tab on a dataset page, but these are limited to the current dataset. With the Chart Builder you have much more possibilities, combining different sources from your domain and from the OpenDataSoft network.
Add a dataset to this layer to start importing data to your chart.
The list of available datasets is displayed: you can search or browse by category to find your datasets. As usual you can preview and filter a dataset before importing it.
As an example, let's take the public dataset "Entreprises immatriculées en 2015" from Data Infogreffe domain.
Filter data to refine the data, but in our case let's just
✓ Select the dataset.
In the configuration, you can choose the appropriate type of graph. Select
By default Y-axis is set to Count, and the plot is a Line. You can choose the type of graph best suited to your data. Try selecting Column chart instead of Line.
Next to this selection menu, you click on the color box to pick another color for the columns.
In the Configuration settings, click on Appearance, then display values on chart. This adds a label with a numerical value on top of each column. Under Y-Axis you can set the scientific display for the Y scale.
Finally, add a threshold color under Color threshold: type *17000* next *Y >* and choose another color. You can add multiple thresholds with different colors.
Here is what this layer looks like with these settings and two thresholds:
Adding a serie¶
You can add a data representation on the same dataset by clicking on
Add a serie.
The X-Axis parameters will be the same, but you can customize the type of curve and tweak the appearance parameters on the new serie.
Try changing the type from
Spline to see a nice curve on top of the initial column chart.
You can reorder the series by clicking and dragging. The series order affects the chart legend and the serie in the top has priority on the Y-Axis parameter.
In the next section we will add another dataset to our chart.
Let's elaborate another data visualization, this time with multiple layers.
We will start the same way as before, click on
Add a dataset to this layer then search for "Oil Prices" to add the Oil Prices dataset.
No need to filter it, just note that the data starts from the year 1987. We will use 1987 as the starting point for the other layers.
When you first import a dataset in Advanced Charts, the default visualization of this dataset is imported: in this case you can see two lines: WTI Spot Price and Brent Spot Price.
Let's add another dataset! Look for "Commodity Prices" to find Commodity Prices since 1980 dataset from the OpenDataSoft public collection.
This time, add the filter "date>1987" so that the X-axis will match with the first layer. Then import the ~20 000 records.
The default visualization is a line with the average commodity price.
You should see your query filter just under the dataset name: "query: date>1987".
You can click on the pencil icon of a layer to change the selected dataset or refine the query.
Under Configuration > Label, set "Commodity Average Price" as the alternate serie title (more accurate than the default title)
This should look like this (you may have different colors):
For the final layer, we are going to use Global Financial Development from WorldBank to see if there are some correlation between those.
There are more than a million records: for the chart to make sense, we will limit the records to the USA. Like the previous layer, we will also keep only records after year 1987.
These conditions translate to this query: "country_code: USA and year>1987`"
As for the second layer, configure a more explicit label name for the legend: "Global Financial Development"
Let's finish with a few last details to polish the Chart appearance
Global Precisionto Year instead of Month so that the legend reflects the data precision.
- Delete WTI Spot Price serie (click on the X cross) and set the X-axis to a Year precision instead of Month
Linein the Global Financi`al Development curve to smoothen the curve a bit.
- Now you can type a name in
Chart titleand click
Save this chartto save create a short URL that you can embed or share!
Here are the resulting parameters:
And the final result of the chart:
Other charts examples¶