Upload

Does my data have to be specifically arranged in order for me to put it into Swivel?

Swivel can handle only one table at a time. If your spreadsheet has multiple tables, copy the table you want to import into a new spreadsheet.

The data in your table must be arranged in columns containing values of similar type, date, if present, must be in a column, and column titles, if present, cannot span lines or columns. If your data isn't like that, you may be able to tweak it in Excel.

Swivel now provides a Swivel Toolbar available for Excel to allow you to manipulate your data into a format that Swivel will be able to understand.

Your data must have a simple table structure. Your data must be in a single table, with values arranged in uniform columns of like items. If there is a date, it must be in a column. If columns have headers, they must be on one row.

Here's an example of a table that will import easily. The date is in a column, text and numeric values are in uniform columns, and column titles are on one line:

However, this data won't import well. The date goes across, the values, totals, and subtotals are all mixed in, and there seem to be categories and subcategories:

You can either copy and paste your data, or upload your data as a csv (comma separated values) file.

Back to Top

Copying and pasting data into Swivel

One way to get data into Swivel is to type in data by hand or paste copied text. If you have spreadsheet program, such as Excel or Google Spreadsheets, available, it's best to create your table in one of those programs. You can then cut and paste the table into the text area.

If you type values directly into the text box, separate values with a comma. If values themselves contain a comma, your best bet is to create the table in a spreadsheet and cut and paste it into the text area. Here's an example of a simple comma-separated table:

Year,Trips
2001,24
2002,23
2003,10
2004,2
2005,31

The data in your table must be arranged in columns containing values of similar type, date, if present, must be in a column, and column titles, if present, cannot span lines or columns. If your data isn't like that, you may be able to tweak it in Excel. However, importing it as is will probably not give satisfactory results. Review data guidelines.

Back to Top

Uploading data as a csv file

You can upload a csv (comma separated values) file directly into Swivel. If you are using Microsoft Excel, select File -> Save As..., and choose CSV (Comma delimited) as the format. If the original spreadsheet has multiple worksheets, make sure the worksheet with the data you want to import is currently selected. On saving, click Yes on all the dialogs presented to you.

Note: Importing a native Microsoft Excel spreadsheet file (with .xls extension) will not work. If your file is already a comma-separated values (CSV) file, you can import it as is.

You can also specify a URL to a CSV file on the Web. Swivel is not smart enough to parse data directly from Web pages Ñ that requires manual steps. See more at uploading data from web pages.

The data in your table must be arranged in columns containing values of similar type, date, if present, must be in a column, and column titles, if present, cannot span lines or columns. If your data isn't like that, you may be able to tweak it in Excel. However, importing it as is will probably not give satisfactory results. Review data guidelines.

Back to Top

Extracting tables from web pages

If the web page table you are trying to upload into Swivel is plain text formatting with nothing else on the page except for a comma or tab delimited table, like this one, you can put the URL for the website directly onto the upload page.

However, Swivel is not yet smart enough to automatically extract tables from web pages. Getting data from web pages into Swivel is tricky, because of the wild variety of formats and structures of data in web pages. Tables in Web pages may have formatting that interferes with data import. Using Excel or another spreadsheet program, you may be able to get the table in the right shape using the following method.

• Select the whole web page that contains the table you want by pressing Ctrl-A, and copy it by pressing Ctrl-C. Paste the copied data into a blank Excel spreadsheet by pressing Ctrl-V.


• Locate the copied table in the spreadsheet (you may need to scroll down or sideways to find it). If you can't find anything that looks like the table you wanted, it is possible that the web page formatting interferes with the structure of the table. Unfortunately, importing such data may not be possible without some manual fixing.
Note: The trick with pasting the whole web page only works on Windows. Excel for the Mac doesn't have this feature. Try copying just the table alone, although this often won't work.


• If you were able to find the table in the spreadsheet, select and copy it. Open a blank Excel spreadsheet and paste the data as plain text by selecting Edit -> Paste Special... and choosing Text. This will remove any web formatting.


• Take a look at your data, now without any web formatting, and check that it conforms to the data requirements for upload: data is arranged in columns, date (if present) is in a column, column titles (if present) are on one line. If columns don't have titles, this is a good time to add them.

If all looks good with your data, you can either copy and paste your data into Swivel, or upload your data as a csv (comma separated values) file.

Back to Top

Supported Date Formats

Swivel currently supports the following list of date formats. If you find one that you need that is not on this list, definitely let us know. (We think it's pretty comprehensive.) The example date/time that is used is Friday, December 15th, 2006 at 12:23:23pm.

	Example	Description
1	December 15, 2006	month_name day, four_digit_year
2	december 15, 2006	month_name day, four_digit_year
3	12/15/06 12:23:23	month/day/two_digit_year time
4	12/15/06	month/day/two_digit_year
5	12/15/2006 12:23:23	month/day/four_digit_year time
6	12/15/2006	month/day/four_digit_year
7	2006/12	four_digit_year/month
8	2006-12	four_digit_year-month
9	12-15-06 12:23:23	month-day-two_digit_year time
10	12-15-06	month-day-two_digit_year
11	12-15-2006 12:23:23	month-day-four_digit_year time
12	12-15-2006	month-day-four_digit_year
13	2006-12-15 12:23:23	four_digit_year-month-day time
14	2006-12-15	four_digit_year-month-day
15	20061215 12:23:23	eight_digit_date time
16	20061215	eight_digit_date
17	2006	four_digit_year
18	FY 2006	fiscal_year four_digit_year
19	06	two_digit_year
20	15-Dec-06 12:23:23	day-month-two_digit_year time
21	15-Dec-06	day-month-two_digit_year
22	15-Dec-2006 12:23:23	day-month-four_digit_year time
23	15-Dec-2006	day-month-four_digit_year
24	15-Dec	day-month
25	Fri Dec 15 2006	day_of_week month day four_digit_year
26	Fri Dec 15 12:23:23 2006	day_of_week month day time four_digit_year
27	December 2006	month four_digit_year
28	Dec 15, 2006	month day, four_digit_year
29	Dec 15, 2006 12:23:23	month day, four_digit_year time
30	Dec 15, 2006 12:23:23 pm	month day, four_digit_year time am/pm
31	Dec 2006	month four_digit_year
32	Dec-2006	month-four_digit_year
33	Dec 06	month two_digit_year
34	Dec-06	month-two_digit_year
35	12/06	month/two_digit_year
36	12/2006	month/four_digit_year
37	Fri Dec 15 12:23:23 PDT 2006	day_of_week month day time time_zone two_digit_year
38	15 Dec 12:23 pm	day month time am/pm
39	'06	'two_digit_year

Back to Top

Swivel Toolbar for Excel

We have made a toolbar for Excel with some handy tools for getting your spreadsheet data into Swivel. It can be used to convert otherwise complicated formatting into something that Swivel can understand. The toolbar is available for Microsoft Excel 2003 and 2007, but unfortunately not for Excel on a Mac.

How it works

Microsoft Excel is a great desktop tool for organizing and manipulating data. But Excel spreadsheets can be formatted in many different ways, and Excel's flexibility often results in spreadsheet data that is hard to upload naturally to Swivel. Here's a typical example, from the OECD:

Cell B9 above contains a typical "data point" that we might like to plot in a Swivel graph: the unemployment rate for native-born Australian males in 1995. But the value in Cell B9 is associated with a lot of other values that are scattered in other cells of the spreadsheet, as highlighted in the screenshot. The Swivel Toolbar adds features to Excel to make it easy to get each data point like Cell B9 lined up with its associated attributes in a separate spreadsheet row. The result? A fully tabular^* CSV file, ready for a one-touch upload to Swivel.

If you are more visually oriented, here is a video about how to use it.

Toolbar buttons

The Swivel Toolbar is split into two groups. On the left are buttons that run custom Excel macros written cooked up at the Swivel confectionary. On the right are shortcut buttons for traditional features in Excel that are handy for preparing data for Swivel.

      (Copy Selection to Scratch Workbook): To use this button, first select a region of cells that you want to work with. (If you don't select a region, the full range of non-empty cells will be auto-selected.) After clicking the button, new Excel workbook will be created, containing a copy of the selected cells. Note: cell formatting is not copied over -- this is often useful, since formatting can often hide data and its placement on the sheet.

      (SuperFill Right): For the currently selected row, empty cells are filled by copying values rightward to the edge of the active region of the sheet.

      (SuperFill Down): For the currently selected column, empty cells are filled by copying values downward to the edge of the active region of the sheet.

      (Convert Matrix to Table): This button is used to copy a matrix (grid) of numbers into a tabular format in a new sheet. For the tabulation to work correctly, the row-headers and column-headers of the matrix need to be specified. To use the button, first select the cells that form the matrix, including the row and column headers as well as the matrix cells. (If you don't select a region, the full range of non-empty cells will be auto-selected.) Then click the button. Two things happen:

          1. A dialog box appears to ask for input in identifying the headers of the matrix

          2. The top row and leftmost column are colored to indicate a first guess at the extent of the row and column headers.

By clicking in any cell in the matrix, the header coloring will be expanded to the edge of that cell. When you have the proper headers identified, click OK. A new sheet will be created containing a tabular representation of the matrix. This new sheet contains one row for each cell of the matrix; the row contains the matrix's row and column headers as well as the cell value.

      (Save Worksheet as CSV): Opens a File -> Save As... dialog for saving the topmost sheet in the current workbook as a CSV file.

      (Upload to swivel.com): Uploads the selected cells to swivel.com, and launches Internet Explorer to allow you to finish the upload. (If you don't select a region, the full range of non-empty cells will be auto-selected.)

The remaining buttons are shortcuts to standard Microsoft Excel features:

      (Fill Right): Shortcut for Edit -> Fill -> Right, which...

      (Fill Down): Shortcut for Edit -> Fill -> Down, which...

      (Delete Row): Shortcut for Edit -> Delete... -> Entire Row, which deletes the entire row of the selected cell.

      (Delete Column): Shortcut for Edit -> Delete... -> Entire Column, which deletes the entire column of the selected cell.

      (Replace): Shortcut for

^* "Tabular" as in a table of data ... each row represents an independent fact, and each column contains values for a single attribute of every fact. Tables (also known as relations) are the basic modeling constructs in relational databases.

Back to Top

What if I don't have Excel? Is there something else I could use to manipulate my data?

There are various different options as alternatives to Microsoft Excel. We encourage you to look into these possibilities if you going to be uploading data into Swivel, especially if often that data is formatted in a complicated way.

Google Spreadsheets provides a spreadsheet program online: all your spreadsheets are stored on your Google account. The functionality is currently pretty limited, but Swivel has one-click integration for Google Spreadsheets, which can come in very handy as this video shows.

If you don't have a Google account and don't want to get one, there are various open source spreadsheet programs that have most of the same functionality as Microsoft Excel does. OpenOffice is an open-source alternative to Microsoft Office, and also includes spreadsheet program that is very similar to Excel: the entire OpenOffice is available for Windows, Mac OS, Linux, and Solaris. Gnumeric is a singular spreadsheet program originally from the GNOME desktop environment, and is available for Windows and Linux.

Back to Top

What the different data types are in Swivel

Swivel allows you to choose the the type of data that you have uploaded. The reason for this is so that Swivel can understand your data more intelligently. Currently, if you don't choose the appropriate data type on upload, you will have to re-upload the data with the correct data types as they cannot not be changed after uploading.

• True or False: Known as "Boolean" values for the more engineering-minded, use this data type if your data column has two values: true and false, yes and no, successful and failed, or something of the like.
• Currency: Use this data type if your data are monetary values; Swivel will include 2 decimal points in any view of this data. Swivel will automatically recognize data that is monetary values if the values have currency characters such as $ or £ out in front.
• Date or Date and Time: This is pretty self explanatory, but any type of date or time data should be labeled as the appropriate data type ("date" for only dates or "date and time" for data including time). This will guarantee that Swivel will create trend graphs with your data. See our supported data formats if you are not sure about whether Swivel will understand your specific date/time formatting.
• Number or Whole Number: If your data is just numbers and doesn't fit into any other category of data, choose this data type. The only difference between the two is that "Number" allows for decimals while "Whole Number" does not.
• Percentage: Use this data type if your data is expressed as a fraction of 100. Swivel will take percentages expressed as 0.348 and multiply them by 100 to get the correct expression of 34.8%. However, Swivel will not understand 34.8% as a percentage: it will also multiply this by 100 if the percentage data type is chosen and make this 3480%, so be careful with this data type.
• Text: If your data column is not a number, but simply textual labels or text, choose this option. Swivel will automatically understand this column of data as a category as opposed to a metric and will only use this column of data as labels for metrics (usually numbers), though this can be changed in data set properties once the data is uploaded.

Swivel also provides the option to not import a certain column of data. Choose the "Do Not Import" option to skip a certain column of data on upload.

Back to Top

Why do I need to give a source for my data?

Data is not only about numbers: it often has meaning in the real-world. Suppose you upload a data set describing the number of carbon dioxide emissions for all countries in 2005. A user looking at this data set may not believe or agree with some of the data, and will want to see the source of the data to verify that in fact these numbers are correct. Without the source, this user may think that the numbers are made up. The source also helps in giving a context to the data: data from a university study and that from a government agency are viewed differently.

We at Swivel take data integrity very seriously. This is the reason Swivel has started the Official Source Program so that our users can see that data uploaded from these sources is credible as well as useful.

Back to Top

What are tags? What's the difference between categories and tags?

Tags provide a way to describe and classify your data with keywords and terms that would be associated with the data. Tags do not form any sort of hierarchy; they are very flexible as you are the one defining them. You can also use tags as psuedo-bookmarks to find the data you are looking for later. Tags are a sort of personal taxonomy that you make up for your data as you see fit. Learn more about tags.

If a tag you want to use is more than one word long, put the entire tag in quotation marks. Otherwise, simply separate the tags you want to use with spaces. For example, you can tag a data set with gasoline "gas prices" or whatever you please so that you can organize your data however you like and others can find this data using the same keywords. You will be able to see these tags on the data set homepage, and you can also hunt for data that is tagged with a specific keyword on the Data page.

Swivel also provides a space for you to categorize your data under a specific topic. The options are Economics, Entertainment, Health, Politics, Science, Society, Sports, Technology, and Miscellaneous. Separate from traditional tags, this allows users interested in a certain topic, say Politics, to find your data based on a certain topic, which can be accessed on the Data page by clicking on a certain category.

Back to Top

Our Company

About Swivel

Swivel's mission is to liberate the world's data and make it useful so new insights can be discovered and shared.

The Internet has been a positive force for the written word, photos, and videos. However, the world's most important data have been completely neglected. Many people live in a world of pure spin and opinion. When people, business leaders and politicians cannot access the facts in an engaging way they'll just ignore them. And when the facts are ignored citizens, communities and investors lose. Without accountability to the facts our world gets worse.

We believe data is most valuable when it's out in the open where everyone can see it, debate it, have fun, and share new insights. Swivel is applying the power of the Web to data so that life gets better.

How to contact us

We are located at 199 Fremont St., 12th Floor, San Francisco CA 94105. In this day and age, though, swinging by our office is probably not the best way to reach us.

If we are online, you can send us an instant message on Yahoo! IM at teamswivel:

or on AIM at teamswivel:

AIM

Or, if you prefer, you can email us your questions to support@swivel.com and any feedback to feedback@swivel.com. Please send all press inquiries to press@swivel.com.

You can get help and tips from members of the Swivel community at the Swivel Help Google Group. You can also ask for data you're looking for on the Swivel Data Google Group.

Interested in working for us?

We are passionate about understanding the world around us and we like to build great stuff (and have fun doing it). Take a look at the job descriptions below. If you're interested in joining our cause, send an email to us at jobs@swivel.com.

We're currently looking for a Ruby on Rails Developer and a Programming Intern.

Ruby on Rails Developer

Swivel is seeking a Ruby on Rails rockstar.

Send an email to jobs@swivel.com.

Our team is currently focused on building what has been described as "YouTube for Data", "Nerdiest Site Ever", "Not Another Social Network", "Mother of All Databases", "Whoa a Scatterplot of My Running History Versus the Weather AND the S&P 500 Index". It's a very cool and addictive website, but we have a huge backlog of ideas that we're tired of prioritizing down for ourselves, a small team of data junkies who heart blocks, enclosures, and syntactic sugar.

We could use your help.

Responsibilities include:

• Implement features fast and ultra-good
• Define new, eye-drooling product features
• Fix obscure bugs using hard-core test driven development
• Repeat, not necessarily in order

Required:

• 1+ year of Ruby (or other scripting language like python, perl, or php)
• 3+ years of web programming
• 2+ years database experience (MySQL preferred)

Ideal Candidate:

• 2+ years UNIX experience (Linux or MacOS preferred)
• HTML, JavaScript (AJAX), CSS skills
• Startup experience
• Agile Methodology
• C patching, hacking, obfuscating

Send an email to jobs@swivel.com.

Programming Internship (Paid!)

Swivel is seeking an enthusiastic senior-level undergrad or graduate student who loves math and statistics and is interested in receiving hands-on experience with data parsing and analysis.

Your typical day will include scouring Web sites, newspapers, and the blogosphere for interesting and controversial topics, writing scripts to gathering supporting data, and then using Swivel to analyze it, discover compelling insights, and share those insights with the world.

You'll work on projects concentrating on discovering and parsing large, real-time data feeds and importing them into our database.

Responsibilities include:

• Work with and improve existing software and develop new tools for data harvesting; these include custom scripts, browser plugins and excel macros to import data to our database
• Help identify and develop methods for working with complex data sets.
• As a primary user, assist Swivel with feedback and product direction.

Ideal Candidate:

• College junior or senior Computer Science major, recent graduate, or graduate student. Must be local to San Francisco Bay Area.
• Proficient with Ruby, Perl, or Python, in particular strong regular expression and web programming knowledge
• Experience with UNIX
• Strong quantitative skills and a love for data and statistics
• Self-starter who is able to work independently
• Passionately curious about the world around us
• Must be able to work 12-15 hours per week (flexible schedule)

Send a resume and cover letter to jobs@swivel.com. In your cover letter, please share with us some matters of interest to you – ideas and issues you would like to explore and gain insights into through Swivel. Please put Programming Internship in the subject line.