| Google Fusion Tables FAQ |
|
|
| Import and data |
What kind of files can I import? How big can they be?
You can upload spreadsheets (.xls or Google Spreadsheets) and csv files. In the current version we allow tables of up to 100MB and we allow up to 250MB per user. When someone shares a table with you it does not count against your quota.
How should I prepare data for Fusion Tables?
Check if you need to prepare your data prior to uploading it into Google Fusion Tables.
If column headers in your spreadsheet span multiple rows, you should edit them so they occupy a single row. Fusion Tables will recognize a single row of column headers.
Some spreadsheets assume the layout will remain constant. You may need to copy a category header into each row to ensure the information will be in the rows as you expect after import. For example, if "Africa" is listed once, followed by records for each country in Africa, you'll want to copy "Africa" into each row.
If computed values are important to you, you'll need to ensure they are part of the data before importing it.
Comments and explanations that are not part of the tabular data should be deleted from the file you upload. You can put these explanations in the description field during import, or use the commenting features in Fusion Tables to annotate your table's data.
Finally, all location information should be combined into one column. For example, if you have address information in multiple fields (street, city, state, zip) or point coordinates (lat, long) in separate columns, use the spreadsheet's CONCATENATE function to bring them into one column, separated by spaces. For step by step tutorial, see Combine Latitude and Longitude columns.
What happens when I choose a header row?
What data types does Fusion Tables support, and can I specify data types?When you import a spreadsheet, the dialog box asks you to specify which row is the header row. The values in this row will be taken as the names of the columns. You can change them by choosing "Modify columns" in the Edit menu. Any rows before the header row will be ignored.
Fusion Tables supports string, integer and decimal values. In addition, Fusion Tables tries to recognize when a column contains values that can be geo-coded or that can be interpreted as date/times.
You can also instruct Fusion Tables about the type of a column by choosing "Modify columns" in the Edit menu.
Fusion Tables doesn't store images. Place the URL pointing to the location of the image on the Internet in a cell in Fusion Tables. The cell will display a preview of the image.
How do I upload large xls, xlsx, or ods files?
We do not support xls, xlsx, and ods files larger than 1MB. But we do support .csv files up to 100MB. We recommend that you export your large spreadsheets (.xls, .xlsx, .ods) to the .csv format and then upload them.
How is Fusion Tables different from spreadsheets?
The goal of Fusion Tables, as with other database systems, is to manage larger amounts of data than spreadsheets typically do. This size difference leads to a focus on a different set of functionalities. For example, Fusion Tables focuses more on bulk operations on the data (filtering, aggregation, merges). These operations are typically not necessary for smaller collections of data stored in spreadsheets. In contrast, spreadsheets preserve complete flexibility in managing data -- you can put any value you want in any cell and work carefully to format your spreadsheet to make it look nice. Fusion Tables requires that the data be tabular with column headers and types for each column.
How do I get credit for my data?
Upon import, you can specify the attribution of the data (for example, "Compiled by John Smith"). The attribution will appear above the table (or any other visualization) right next to the table name. If someone merges your data with other data, the attribution will still appear along with any other relevant attributions.
What about Intellectual Property issues?
Some of the data you can access has certain Intellectual Property (IP) restrictions. The most prominent example is data that is governed by the Gnu Free Documentation License, such as data from Wikipedia. In a nutshell, the license states that when you make additions to the data, you need to make them publicly available, if you're sharing the data with people outside your organization. You will see a notification about this when you perform certain operations on the data. Don't believe our explanation. Go and read the license description if you think this applies to you.
Separately, we'll respond to clear notices of alleged copyright infringement. For more information or to file a DMCA request, please read our copyright procedures.
Can I disable exports of my data?
Some data owners may not want others to export their data out of Fusion Tables into the CSV format. On import, you can uncheck the box next to "Allow viewers to export." This also has some consequences for exports of derived tables. Suppose user A uploaded a table and designated it as not exportable. Then user B merged A's table with her own table. User B will not be able to export the merged table. Of course, user B can still export her own columns by exporting her own original table.
We make periodic backups of the data, but we don't typically save versions of the data that are earlier than our latest backup. If you need more extensive backups, you can export your data into .csv as needed.
There are several ways to delete data in Fusion Tables:
- To delete a row, click on the trash icon at the right end of the row. To erase a value in a cell, overwrite it or click into the cell and use the delete key on your keyboard.
- Comments that you make and comments on data that you own will show a "delete" link. Click to delete a comment.
- To delete a table, click the "x" next to it in the list of tables. Deleted tables go into the trash, where they are hidden and do not count in your quota. Things get complicated when you consider tables that were derived from the deleted table. Suppose tables A and B were merged to create table C. Now suppose table A is deleted by its owner. What happens to table C? Table C continues to have the values from table A that were present up to the deletion point. Table C does not lose the columns it obtained from table A.
| Types of tables |
To kick off the fun, we've put a few tables in a gallery for everyone's use. You can find these tables by clicking on "Table Gallery" on the landing page. You can merge these tables with your own data if you wish.
Some of the data sets made public by Fusion Tables users have also been added to the Table Gallery. Contact us if you'd like your table included.
Some tables have a different icon next to them in the table listing. Why?
There are two types of tables. The tables you import or someone else imports are called base tables.
Tables that are created from other tables are called derived tables and are shown in the table listing with an icon that includes a plus sign. You make derived tables by creating a view that includes a subset of the columns of an existing table, or by merging two or more tables. There are some restrictions on the modifications that can be made on derived tables. For example, column types can only be changed in base tables.
| Exploring the data |
What kind of queries can I issue?
You can filter and aggregate your data from any visualization. To begin, click on "Show options (filter/aggregate/choose columns)" above the table or visualization. You will see three tabs for filtering, aggregation and column selection.
In the filter tab, select one of the attributes in the table from the drop-down menu and a comparator (such as =, <,>, <=, >=, "starts with" , "ends with", "contains", "does not contain" , "not equal to" or "matches" ), and enter a value. Then click Filter. With the exception of "contains case insensitive," filters are case-sensitive and will match exactly on the text capitalization. Only the values that satisfy the filter will be shown. (For SQL folks, this is like a predicate in the WHERE clause.) You can add more filters, and the result is the conjunction: only rows that satisfy all filters will be in the result.
In the aggregate tab, you have two choices to make:
First, choose a column (or set of columns) by which to group the data. For example, suppose you have a table with the population of each county in each state. You may want to group all the rows that have the same value for "state" in order to compute the total population in each state. Check the radio buttons next to the columns to select them. ("state" in this example). Next select the aggregates to compute by checking the appropriate checkboxes next to each attribute. Aggregation operators only work on numeric columns. A count of the number of rows in each group will always appear in the result, so there is no need to specify it explicitly. If you don't select any grouping columns, nothing interesting will happen.
- which columns to group by, and
- which aggregates to compute for each group.
Can I combine Filter and Aggregate?"Matches" is based on SQL's LIKE, and therefore provides some basic pattern matching functionality. Use the underscore _ for a one character wildcard, and % for a wildcard that matches any number of characters. For example, Name matches Reb% would find:
Rebecca
Rebeca
Rebekah
Rebbekah
You can combine filtering and aggregation to a limited extent right now. If you apply a filter and then click on the aggregate tab, you will be aggregating only the rows that satisfied the filter. If you perform an aggregation and then click on the filter tab, the new filter you create will apply to the rows of the table on which the aggregation was performed, but you cannot set filters on the aggregation columns. For SQL fans, this means we do not yet support the HAVING clause, but predicates in the WHERE clause can be specified even after aggregation.
What happens when I choose columns?
"Choose columns," the third tab under "Show options," allows you to create a view of the data that contains a subset of the columns in the table. Suppose you've created a table that has the details on employees. You want to share the data with collaborators, but you cannot share all the details (such as the salary column). To do this, select a subset of the columns and give this view a different name. You can then share the view with your collaborators. The view will stay in sync with the table from which it was created: any relevant updates that are made to the original table will be seen in the view as well.
Why is there a difference in the background colors of some columns?
The different colors of columns illustrate the original sources of the data. When a table is created by merging multiple underlying tables, it is often useful to know which source each column came from. You can see the legend for the colors in the "About this table" option in the File menu.
| Visualizing data in charts and maps |
Sometimes there are more effective ways to look at your data than rows in a table. For example, when your data contains geographical locations, viewing the data on a map or as an intensity map (where each country is colored according to some attributes) may reveal patterns in the data that are otherwise very hard to see. Fusion Tables offers several visualizations that will help you explore your data more effectively. You will see them in the "Visualize" menu.
Why are some visualizations not available on my data?
Some visualizations require the data to have special columns. For example, you cannot display data on a map unless you have a column with geographic locations. Fusion Tables tries to identify which visualizations are relevant for a given data set, and even give you a default mapping from your columns to the columns required by the visualization.
Visualizations that cannot be displayed are grayed out. If you think the system is missing some visualization opportunities, check on the data types of your columns (by choosing "Modify columns" in the Edit menu). You can instruct the system that a column has a particular data type.
How can I embed visualizations in other web properties?
Some of the power of visualizations is the ability to embed them in other web properties. For example, if you have a blog about the fresh water crisis, you may want to embed a visualization that shows access to fresh water in different countries in the world, like Circle of Blue has done.
The embedded visualization represents public access to your data. You need to make the table public in order for the embedded visualization to work.
Most visualizations show a button "Get embeddable code". Click this button to get a piece of HTML that you can include in almost any other page.
To embed a visualization in Google Sites, see the step by step tutorial Embed in Sites.
To embed the Google Maps visualization,
- under Visualize > Map, click "Get KML Network Link".
- Go to maps.google.com, and paste the link in the search box. Click Search Maps.
- Click Link in the blue bar to get a URL or the HTML embed code.
One of the markers on the map is in a wrong location. How do I fix it?
Most likely, that one has geocoded incorrectly. It isn't always obvious where to put a location on a map. Did you know that there are 70 cities in the United States alone named Springfield?
Google Fusion Tables offers a geocoding correction feature so you can tell it the correct interpretation of the location to use. When viewing the data as a table, move the mouse pointer over the cell with the location name. Click the globe icon that appears. In the pop-up window, search with any text until you see a map pin located in the correct place. Click to select that pin, and Google Fusion Tables will now use the new pin's location.
Note that if you just edit the text in the table view, the location on the map may not be re-computed.
How do I specify which column has location information?
Fusion Tables will automatically detect in most cases when a column contains information that specifies a location, such as a latitude-longitude pair or an address.
To manually specify which column Fusion Tables should use as a location for map visualizations, choose Edit> Modify Columns. Click on the column in the list, and then specify type "Location" in the column type menu.
To ensure Google Fusion Tables geocodes the location as correctly as possible, put all location information into one column, separated with a space. For example, put latitude and longitude into the same column in that order, separated by a space. Or, put all parts of an address including street, city, state and zip into the same column, separated by a space.
Tip: many spreadsheets have a "concatenate" function to help you combine information from multiple columns easily into one. For an example, see Combine Latitude and Longitude columns.
Tip: use a text editor's find-replace function to remove the commas between two values.
What geo standards does Google Fusion Tables use?
The Intensity Map visualization uses a controlled vocabulary of country names and 2-letter country codes, as specified by the ISO 3166 standard. See the Wikipedia reference page. http://en.wikipedia.org/wiki/ISO_3166-1_alpha-2#Officially_assigned_code_elements
Fusion Tables assumes that latitude and longitude numbers are in Decimal Degrees. Lat-long info in Degree-Minutes-Seconds should be converted to Decimal Degrees.
Like all other Google mapping tools, Google Fusion Tables assumes the WGS 84 datum. If you know your location data is in a different datum, you'll want to convert it first for the most accurate mapping.
How do I export to KML for Google Earth?
Click the "Export to KML" link on Visualize > Map to create a KML file from the current set of rows.
The Map visualization shows 200 rows at a time from the found set. If your current found set is larger than 200 rows, not all will appear on the map but all will be in the KML file.
If a filter is currently applied, only that filtered set of rows will be included in the exported KML.
How do I specify what information appears in the map pin balloons?
Under Visualize > Map, click "Configure Info Window" and select the template to use. The default template lists all the column names and their values for that map pin.
Other templates are designed to show photos, links and more. Select which column's values should be used to populate the parts of an information balloon template. For example, for "photo link" use a column where each row has a link to a photo in order to include those photos in the map balloon.
When the "Export to KML" link is clicked, the current configuration of the info window will apply for the exported data.
How do I show "live" Fusion Tables data in Google Earth?
Use the "Get KML Network link" feature. For a step-by-step tutorial, see Live update map from Fusion Tables or large spreadsheets.
Note that like any embedded visualization, the table must be made public for KML network link to work.
| Sharing and merging |
How does sharing work with Fusion Tables?
Who can I share my tables with?Google Fusion Tables offers several options to share data with others, just like Google Docs but with an additional wrinkle. You can invite people to share your data in the following ways:
Owners can send invitations from the sharing dialog accessed by the Share button at the top right corner of your table. Users who are not owners can use the Permissions button to see the other people with whom the table is shared.
- As viewers: In addition to viewing your data, users with viewer permissions can merge your data with theirs. However, they can only add your columns to their own tables; they can't add their columns to your tables. Viewers can also comment on the data.
- As contributors: Contributors can add columns to the tables you share with them, as well as having all viewer privileges. You can invite contributors only on derived tables.
- As collaborators: Collaborators can modify the data in your table, as well as merging and commenting.
- As owners: Only owners can invite others to view, contribute, or collaborate on a table.
Can I share only a part of my table?You can share your tables with any email address.
In order to access a table with the permissions you've granted, the invited email address has to be activated as a Google Account. Any Gmail address is automatically a Google Account, as is any email address you've used with other Google products such as Groups, Picasa, AdWords, or Checkout.
In addition, you can set up a Google Account with *any* email address -- look for the "Create an account now" link on the Fusion Tables sign-in page. You can find more information on Google Accounts at Accounts Help.
Can I make my table public?By combining the creation of views with sharing privileges, you can define permissions for subsets of columns. For example, Adam can create table A and then make a view, called table B, that includes columns 1, 2, and 3 of A. Adam can invite Beth to collaborate on table B. In the meantime, Adam can also create a view called table C that has columns 1, 3, and 4 of table A. He can invite Carina to collaborate on table C. So Beth and Carina can both see columns 1 and 3, but each one has access to a column that is not seen by the other.
How do I ensure people see the data the same way I'm looking at it?To make a table public, check the box next to "Anyone may view this table" at the bottom left of the sharing dialog.
When this box is checked, anyone with the URL to your table will have "viewer" access. Your table may also be indexed for search engines and listed in Fusion Tables public tables list.
Because a visualization of your data embedded on a blog or other external website implies you are comfortable with the values in the table being publicly viewed, this box must be checked to activate the "Get embeddable code" features.
When you've applied the filtering, aggregation and visualization that you like to the data, click the "Snap It" button for a URL that will show the data in this same configuration. You can send this URL by e-mail or IM, or link to it from a website. The access controls you set up for your data will still apply.
How do I share some of my columns with others, so that they can merge their data with mine?
To enable others to merge their columns with yours, proceed as follows:
- Use "Show options > Choose columns" to create a view of your table that includes only the columns you want to share. Select the columns you want to share, then enter a name for this view and click Save View.
- Share the new view with your collaborators: Click Share in the top menu bar (far right), check the "contributors" option, and specify the email addresses of the people you want to share with. They will then be able to merge their columns into this table view.
Collaborators don't always agree on data, and sometimes discussion is needed to understand the assumptions underlying the data, to learn how it was collected, or to reach an agreed-upon correct value. With Fusion Tables you can discuss data at multiple levels of granularity: rows, columns, and individual cells.
To initiate a discussion, click on the chat-like icon that appears when you hover over a cell or a column name. To initiate discussions on rows, click on the chat icon on the right end of the row. Enter your comment and click Save. Cells that have comments are identified with a small triangle at the upper left-hand corner. Note that if someone changes the value in a cell, the change will appear as part of the discussion trail.
The important thing to remember about discussions is that they are associated with individual tables. As a consequence, you can have several disjoint discussions on the same data! Consider the following example: Suppose table A was merged with table B to create table AB. In parallel, table A was merged with table C to create table AC. The discussions on AB and AC are disjoint from each other, even if some of them are on data supplied by table A. In fact, the discussions on tables AB or AC will not appear when you look at tables A, B, or C.
How (and why) do I merge tables?
One of the goals of Fusion Tables is to enable the combining of data from multiple sources and people -- that is, the merging of different tables. To merge two tables, you'll need to identify a column in each table to use for matching up values. (For SQL fans, this is a join on primary key.) Since the values in these columns will identify the rows for matching, you'll need to make sure that within each table, no two rows have the same value in that column. To merge data with the table that you are currently viewing (we'll call it table A), you proceed as follows:
- Click on Merge in the menu bar.
- Select the merge column in table A.
- Select a table to merge with (we'll call this one table B). You can use the drop-down menu to select one of your own tables or a table in the Table Gallery, or paste in the URL of a table that someone shared with you.
- Select a merge column in table B.
- (optional) You can specify that only a subset of the columns from each table be included in the merge result. To select columns, first check the "Select subset of columns" box below the table, and then check the columns you want to include.
- Give a name to the resulting table. If the table you're starting with is already a derived table, then you don't need to create a new table; you can simply merge the new columns into the existing table.
Why are rows dropped in the merge?
When you merge two tables, there may be values that appear in the merge column of table A but not of table B, and vice versa. In the current version of Fusion Tables, table A (the table on the left, which you started with) dictates what rows appear in the merged table. (Database buffs will recognize this as a left-outer join.) For example, suppose table A has a row where the value of the merge column is "Argentina," but table B doesn't have a row for Argentina. The merged table will have a row for Argentina, but only with the columns from A (the rest of the columns will be empty). On the other hand, suppose table B has a row for Vietnam but table A does not. In that case, the merged table will not have a row for Vietnam. If this is not your desired result, try reversing the roles of A and B (that is, starting the merge from B).
How do merged tables correspond to their constituent tables?
If table C was created by merging tables A and B, then any update to A or B will be reflected in C. If you have edit permissions to a column in table C, then any edits you make will be reflected also in table A or B.
How can I use publicly shared data tables in my own data set?
When you're viewing a public data set, you can combine it with your own data by choosing Merge. You will be prompted to choose a name for the table you created. Hence, you won't be modifying the public data, just creating your own view that includes it.
You can make one of your own tables public by checking "Anyone may view this table" in the Share dialog box. Your public table will not be part of the gallery, but you can send the URL around to anyone you want to share it with. In the future, Fusion Tables will support a search option for public tables.
| Programmatic Access |
Does Fusion Tables have an API?
Yes. Check out the documentation at Google Fusion Tables API .
| Contacting the team |
Where can I send my feedback or ask questions?
You can send us feedback at googletables-feedback or post questions on the fusion tables user group .