IMPLAN Data Library gives users the ability to view IMPLAN data in one of three ways: Explores, Looks, and Dashboards. These are essentially the same data points packaged in different ways for different use cases and Data Library proficiency levels. Explores are where users build and run data queries in Looker by filtering, selecting, and pivoting Dimensions and Measures. Looks are saved reports or results of a data query. Dashboards are a collection of tiles or data query visualizations (Explores & Looks) with an answer to a particular question.
This article will walk through each of these options and considerations for utilizing them.
STARTING FROM SCRATCH
An Explore is where users build and run data queries in Looker by filtering, selecting, and pivoting Dimensions and Measures. There is a designated area for Explore in Looker, but there are several other access points.
Start building queries from scratch by selecting a Data Library Asset from the Explore dropdown menu.
Each Data Asset has Dimension and Measure options to build data queries.
- A Query is a question using data, displayed in Looker as charts/tables and visualizations.
- A Dimension is a qualifying field of a data query or report. These are characteristics of the data that will be displayed in the query (examples: Region, Industry).
- A Measure is a quantitative field of a data query or report. These are the data points that will be displayed in the query (Total Employment, Total Value Added, Kilograms of Greenhouse Gases).
Filtering by a Dimension or Measure limits the data displayed in the query based on the rule(s) set.
Selecting a Dimension or a Measure will add the selection as a column in the data query. Click Run in the upper right corner of the screen to populate and update the query after making changes in any of the fields. Increasing the Dimension and Measures in a data query will increase the expected load time. Load times for data queries may be longer in certain Data Library Sections than others, such as Trade Flow Data and Core Competency Data due to the amount of data contained in these sections.
In certain Data Library Sections Filter-Only Fields is a third field in addition to Dimensions and Measures.
The possible Filter-Only Fields include:
- Aggregation Scheme Param: choose to view data at the most granular level, or view data at a more summarized level by selecting to group Industry/Commodity Dimensions by 2-Digit or 3-Digit NAICS.
- Dollar Type: Compare data in like dollar years by selecting Real. Unadjusted dollar values can be observed by selecting Nominal.
- Region Type: Select the level of regionality of data to include in the query.
- Field Picker: The Field Picker allows users to select a Measure from a dropdown menu. In order to include and/or Filter by the Measure selected in the Field Picker users need to select the Measure titled Metric.
FILTER DIMENSIONS & MEASURES
Applying Filters to the appropriate Dimensions is critical to ensure unintended data is not included in the data query. For example, when no filter, selection, or pivot is applied to Data Year, then the selected Measures displayed would be summed across all Data Years in the model. Filter by a single year to view the Measure for just the desired year.
Filtering by a Measure is less common, but can be helpful when setting a threshold on a Measure is appropriate. For example, if I only wanted to see all Industries that employ over 100 people I would filter my Employment Measure and select Greater Than in the Filter window from the dropdown menu and type 100 in the field and click run.
SELECT DIMENSIONS & MEASURES
Selected Dimensions break out Measures by rows. For example, if State is the selected Dimension and Total Value Added is the selected Measure, clicking run will produce a table of 51 rows (one for each State and DC), and two columns one for State name and one for Total Value Added. This table displays Value Added by State. Clicking on the Value Added column heading would sort the table in descending order (carrot in column heading point down). Clicking a second time would sort in ascending order (carrot in column heading point up). This is just one simple data query that can be produced using the Industry Data. Selecting an additional Measure would add another column of data points. Selecting Total Employment would add a column displaying Total Employment by State in addition to Total Value Added.
If the Data Year Dimension was selected instead of filtered, then Data Year would be included as a column in the table and Measures would be further broken out by year. Selecting Data Year in the Value Added by State query example would add a Data Year column, and the table would then include a row for each State and Year combination. By default the data query has a row limit of 500, so when that limit has been reached, the following message will display, “Row limit reached. Results may be incomplete”. In the data window users can update the Row limit value to up to 5,000.
When downloading data, there is a row limit of 100,000 (Click the Gear icon in the upper right corner, select Download, then toggle Limit option All Results). There can be a maximum of 200 columns.
Sort by multiple columns (sort by one column then another just like in Excel) by sorting a column, then holding Shift + sorting the next desired column. In the Value Added by State and Year example, the table could be sorted by State then Year, Year then State, or even Year and then Value Added to view States ranked by GDP each Year, and so on.
WHEN TO FILTER & SELECT
In some cases users may want to filter and select a given Dimension, particularly when Filtering by several Dimensions. For example, if my research is focused on just NC, SC, and GA, I could filter for these states. Applying this filter without selecting the Dimension would cause each Measure to be a sum for the three states (showing just one data point for each Year). Filtering and selecting on State would include a State column and produce rows for each of the three States.
Pivoting a Dimension will convert the options of the given Dimension into columns, transposing the data. For example, pivoting by Data Set Year will create a column for each Data Set Year so that data can be compared over time side by side, instead of row by row. With this setup in the Value Added by State example, I could compare Value Added over time in each State row. Alternatively, pivoting on State would allow me to compare it across regions in each Year row.
VISUALIZATIONS & OTHER OPTIONS
Query results will appear in two windows below the Filter window, a Visualization window and a Data window.
Updates like sorting in the Data window will affect the Data displayed in the Visualization window. Toggle the Totals box to add a column for each row total. Check the Columns Totals box to add a row at the bottom of the table for each column.
Clicking the Add Calculation button allows users to add calculations into the query, such as calculating the difference between columns.
Find more information from Looker:
- Exploring Data in Looker
- Using Table Calculations
- Creating Visualizations and Graphs
- Visualization Types
Looks are saved data queries or reports. Saved Looks contain a Filter and Visualization window (mirroring the format in Explore).
When in Explore, the Explore Actions icon (gear) in the upper right gives the option to Save as a Look. Selecting this option saves all of the chosen filters, visualizations, fields, sorting, and so forth from the Explore. The saved Looks can be added to multiple Dashboards as Look-Linked Tiles.
Looks can be deleted from within the personal folder by clicking the Tile actions for the Look and selecting Move to Trash.
Dashboards are collections of Explore and Look-Linked Tiles on the same page. IMPLAN has pre-built Dashboards that are a helpful starting point for each Data Library Section. Find these under Dashboards along the top of the page within the Data Library.
Learn more about each of the respective Dashboards here.
CREATING A DASHBOARD
Users can also build their own Dashboard in the Data Library! Creating a Dashboard allows users to pull together visualizations of multiple data queries and add text to build a research summary. Dashboards are made up of Query Tiles, Look-Linked Tiles, and Text Tiles. Text Tiles include text, which is helpful for adding headers, notes, and other descriptive text to complete the research summary. Note that the selected filters will be applied to all tiles within the Dashboard.
Query Tiles can be defined from an Explore by saving a Report to a Dashboard. Use the Explore Actions icon in the upper right for this option.
There are several routes to starting a query visualization that will be added to a Dashboard. Data can be explored and added by starting from scratch with an Explore, or by exploring a Look or a tile of an existing Dashboard.
Looker provides additional resources to help users Create and Edit Dashboards and Reports (Looks). Checkout the video and see the Creating and Editing User-Defined Dashboards section!
User Defined Dashboards will be saved under Dashboards > User Defined Dashboards.
A User Defined Dashboard can be deleted from within the Dashboard by clicking the Tile actions (three dots in the upper-right corner), then selecting Move to trash.
EXPLORE FROM DASHBOARDS AND LOOKS
An Explore can also be started from an existing Look or a tile of an existing Dashboard. Click on the Tile actions (three dots) on any given Dashboard tile selecting Explore From Here.
Click Explore From Here at the top of the page to Explore from a Look. This will open the equivalent Explore that it would require to create this tile within a Dashboard or Look.
Written August 30, 2023