Google Docs Chart

I started with the default Google Docs pie chart. (In a Google Docs spreadsheet, Insert > Chart.) It has relatively few options and no control over the colors, font or type size. In a spreadsheet, the chart is adequate, but the text is rendered with a different font that I don’t like. Conclusion: Passable but not desirable.

Google Docs Gadgets

Next, I looked at the three Google Docs pie chart gadgets (In a Google Docs spreadsheet, Insert > Gadget.) which were initially promising. One of the pie chart gadgets looks functionally identical to the previous one, but when it’s rendered as an image it has a frame and a Google Docs logo. It renders with the correct font, which is good, but I didn’t want the frame. Conclusion: Keep trying.

There’s a second pie chart gadget by Google that has similar controls but renders with different colors and in a different type size, along with a border like the previous one. To my eye, the colors don’t go well together and the border feels too tight — so tight that there’s no white space between the legend and the border. Finally, the title doesn’t render as part of the image. Conclusion: Hell, no.

Finally, there’s the FusionCharts Pie & Donut Widget by InfoSoft Global Ltd., which has many more controls than the other two, including font and type size, but it is perhaps the ugliest of the three, with a beveled circumference of the pie and a shaded background, poorly centered in a frame. There is a palette control, but that appears to affect the background and text color only, which is disappointing. Most annoyingly, displaying the editing panel for this gadget expands the rendering area, potentially resulting in an ultra-tall frame if you’re not careful. Conclusion: This is the ugliest yet.

Google Chart API

Finally, I decided to give the Google Chart API a shot. Unlike the Google Docs charts and gadgets, the API has no user interface, so you have to learn the syntax in order to use it, but I thought it was worth the effort to create something that was exactly what I wanted. Well, close enough to what I wanted. I’m still not thrilled with the results, but it was good enough to suit my purposes.

It’s not for the faint of heart, though, since you have to learn the syntax to pass the right parameters to the server in the URL, so this:

http://chart.apis.google.com/chart?cht=p&chs=500×240&chtt=Posts+by+Category+-+Chart+API&chd=t:30,23,22,14,12,5&chl=Reputation|Identity|Influence|Meta|Tips|Miscellaneous

Becomes:

I was generally pleased with the results, though the type is a little small for my taste. The colors match the look here, but that’s purely coincidental.

The syntax is well documented, but I don’t understand why there are different delimiters for each parameter. Otherwise, it’s easy enough to figure out, if you’re patient and don’t mind reading documentation. Here’s how a simple chart URL breaks down.

Conclusion

In the end, I was disappointed with all of the choices. I liked the flexibility of the Google Chart API, but it takes a lot of effort to get the results you want. I was underwhelmed with all of the easier chart options built into Google Docs, and I was even more disappointed when I learned that you can’t embed a chart from a spreadsheet into a word processing document. I appreciate the simplicity of Google Docs, but the chart tools have a long way to go.

The hard way

Here’s some of the sample data I used in the previous conversion example. I wanted to split the Twitter ID, first name and last name into individual columns.

There are several common spreadsheet functions that can be used to coerce data from one form to another. Used together, they can be extremely powerful, albeit inelegant. The key is to break down complex problems into many simple problems and combine the solutions.

You can write complex formulas with functions nested within functions, or you can temporarily add extra columns to your spreadsheet and do things step by step. For the sake of illustration, I’ll go one step at a time.

Step 1: Cleanup

The first thing to get out of the way is data cleanup. Extra spaces where you don’t need or want them can lead to unpredictable results. Take a look at cell A2.

You can’t see them, but there are four spaces before the asterisk and an extra space at the end of the line. I used the trim(text) function to strip any spaces from the beginning and end of the cell so I have clean data to work with.

To the naked eye, the result looks the same, but the leading and trailing spaces are gone. If I knew that there was a consistent number of spaces in every row, I could skip this step. I decided to play it safe.

Next, I wanted to get rid of the asterisk and the space at the beginning of the cell. The functions in Excel and Google Docs spreadsheets are all about getting the results you want, not eliminating the results you don’t want, so you have to think about the problem in that light.

There is no function to remove two characters from the beginning or end of a cell, but there is a right(text,number) function to get characters from the end of a cell. (Similarly, there’s a left(text,number) function to get characters from the beginning of a cell.)

The challenge is that you have to tell the left() and right() functions how many characters you want. In this case, I wanted all but the first two, so I had to use the len(text) function to find out how many characters are in each cell and subtract two (the number of characters I didn’t want).

Alternately, I could have nested those two steps and achieved the same result.

Step 2: Extracting the Twitter ID

The next step is a little trickier. I wanted to extract the Twitter ID, which varies in length, so I had to grab the correct number of characters each time. I know that Twitter IDs are all one word, and there’s a space after each ID, so I used the find(find_text,text) function to locate the first space character in the cell and grabbed everything to the left of that using the left() function. I had to put the space in quotes for it to be recognized as a literal text string.**

Again, I could have nested those two steps and achieved the same result.

In fact, I could have nested the cleanup and the text extraction steps in one long formula: =left(right(trim(A2),len(trim(A2))-2),find(” “,right(trim(A2),len(trim(A2))-2)))

It looks ugly, but if you build it step by step from the inside out, it’s not much more difficult than the alternative.

Step 3: Variations on a theme

It’s clear sailing from here. Mix and match left(), right(), len() and find() to extract the first and last names — same methods in different combinations.

Once I found the Twitter ID, I was able to find the full name by extracting the characters to the right of the Twitter ID. I also used trim() to get rid of extra spaces.

To find the first name, I extracted everything to the left of the first space in the full name…

…and I extracted everything to the right of the first space to find the last name.

Take it easy

Nested formulas can look and feel intimidating, even to spreadsheet pros. Sometimes it’s easier to create temporary columns rather than nesting formulas. You can see the results of your work and identify problems as you go.

Once everything is working, you can either make a nested formula by copying things that work into new formulas, or you can hide the columns with your intermediary steps so your spreadsheet looks neat and friendly.

A well-designed spreadsheet can make your lists more useful by automating tedious tasks. Organization is just the beginning.

1. * Keep your document around and you’ll have a formula that you can paste into other spreadsheets too. Good work deserves to be re-used.
2. ** As opposed to looking for the contents of one cell in a longer cell, which is useful but doesn’t make sense in this case.

If you already have semi-structured information that you’ve created in a word processor or copied from a web page, it can seem like too much of a pain to try to wrangle it into a spreadsheet. But if you ever need to sort or filter your data, perform calculations or create reports, a spreadsheet is a better format for your data.

There are two ways to turn semi-structured text into neat rows and columns: the easy way and the hard way.

The easy way

If you have neatly organized and structured text, like comma-separated values, most spreadsheet applications have an import function to can convert text to columns. In this example, however, I just copied text from a web page and pasted it into a blank spreadsheet. The information consists of Twitter ID, first name and last name, and I’d like each to appear in its own column.

If you use Excel 2007, the Text to Columns feature makes it easy to turn text like this into columns.*

The Text to Columns feature in the Data tab of the ribbon opens the Convert Text to Columns wizard. It can’t get much easier than this.

Step 1: Delimited text

Unless you’ve laboriously organized your information so that each column takes up exactly the same number of spaces (yeah, I didn’t think so) you can just click Next on the first screen.

Step 2: Set the delimiters

This screen lets you identify which characters distinguish one column from the next. The cool thing about the wizard is that you can choose several delimiters if necessary. Tab is selected by default, but my sample data contained no tabs, so you can see in the Data Preview pane that the text won’t be split into distinct columns.

The hyphen is a consistent delimiter in this text, so checking Other and typing a hyphen (-) in the field next to it tells Excel 2007 to use that character as a column delimiter. Now you can see in the Data Preview pane that it’s getting closer to what I want.

Depending on your data, you may want to try using a space as a delimiter. In many cases, such as with addresses, the results could be disastrous, but my sample data is very simple, so the benefit of turning the first and last names into columns outweighs the problem with the compound last name in the third line of the preview. Since this is a small data set, I can easily find and fix that one problem once the text is converted.

Step 3: Choose columns

Finally, the last page of the wizard lets you set the data format for each column. This is important if you want Excel to recognize dates correctly, but in my case, since I didn’t have any date columns, it wasn’t necessary. Finally, I selected the first two columns and clicked Do not import column (skip), since they contained no useful data.

The result

The result is almost exactly what I wanted. There is a problem with one of the rows, but it stands out enough that I can fix it with minimal effort.

I used a very small data set for this example, but this would work just as well with a much larger data set, and the time required to fix little problems like the one above would be significantly less than it would have taken to copy and paste the information into individual rows and columns.

So that’s the easy way. The hard way is next.

1. * I don’t remember if it Excel 2003 had a similar feature or where it might be.

Let’s say you have a spreadsheet that looks something like this…

You can combine the contents of different cells to arrange the information however you need it or have the contents of one cell display in many places so you only have to update it once.*

Formulas

All the cool spreadsheet magic starts with a formula, which can be big and nasty and complicated but can also be very simple. Start by typing an equals sign (=) and then the formula you want to have calculated. It can be a mathematical formula or it can be a reference to the contents of another cell, which is identified by the cell’s column and row. For example, the cell in first row of column A is cell A1.

You can combine the two forms, so you could multiply a cell by another cell, for example. You can even refer to cells on other worksheets by referring to the worksheet in single quotes (’) followed by an exclamation point (!) and a cell reference.

Concatenation

The ampersand (&) character lets you concatenate (i.e. join) the contents of two cells together.

By itself, that isn’t necessarily useful because the cell contents are stuck together without any space between them, but in conjunction with literal text, you can do some very cool things.

Literals

You can put text inside quotation marks (") to display that exact text. You can concatenate cells with literal text to organize information in as many ways as you want. Practically speaking, this means you can have one range of cells or worksheet to hold your data and other worksheets for different arrangements of that data.

It looks worse than it really is. The first example above might look intimidating but it’s really fairly simple: the contents of cell A2, space, the contents of cell B2, space and open parenthesis, the contents of cell C2, close parenthesis. It’s really just about organizing information.

Creating useful, flexible lists and reports takes a bit more thought, but just about everything else you need to do in a simple list of data is an application of these basic spreadsheet skills.

Three kinds of punctuation. That’s it. Give it a shot.

1. * You can do math too — the original purpose of spreadsheets — but I hear a lot of PR pros don’t like math.

Given how often we use spreadsheets, you’d think that PR people as a group would have some mad ninja Excel skillz, but that doesn’t seem to be the case.** We tend to use it as list-making software.

Because of that, many of the spreadsheets that I see are brittle, inflexible and heavily formatted for a particular display or output format. How often have you paid someone at your agency (or billed your client) a couple of hundred bucks an hour to copy, paste and reorganize rows in a spreadsheet? If the answer is greater than zero, you’re wasting money.

Here are some simple tips for making spreadsheets more flexible.

One thing at a time

You can save yourself a lot of time, money and frustration by organizing your data efficiently. Make sure each column contains only one instance of only one kind of information. Database geeks call this ‘normalization’.***

Rather than have a column contain contact information like this…

…you want something like this:

The payoff

Keeping the information in its most atomic form with one type of data in each column lets you sort and filter by any of those columns, which makes it easier to work with large tables of information.

• Sorting makes it easier to find the information you’re looking for and helps you identify redundancies or anomalies in the data.
• Filtering out information you don’t immediately need helps you focus on what’s relevant.

Finally, it’s a lot easier to combine spreadsheet columns in new and interesting ways than it is to wrestle them them apart. If you have an existing spreadsheet with different kinds of data glommed together in one column, you can break it into columns without too much pain. That’s a topic of an upcoming post.

Put a little thought into the design of your spreadsheets and you’ll have better insight into your data, more flexible documents, and a little less aggravation with your spreadsheets. What more could you ask for?

1. * For basic spreadsheets, Google Docs is an excellent no-cost alternative to Excel.
2. ** I once worked with someone who used Excel every day but didn’t know it can do math.
3. *** True database normalization involves a lot more, but this is a basic principle.