1
$\begingroup$

I have a sample of records, with string and numerical columns. My sample is currently hosted in an excel spreadsheet. I need a tool that wil produce discripte statistics for each colomn, such as max and min values, number of unique values, max string lenght. Are there any such tools available. Any tips? The ultimate goal is to derive a data dictionary.

$\endgroup$
  • 1
    $\begingroup$ Spreadsheets are not good models for reproducible research. $\endgroup$ – Frank Harrell May 3 '13 at 12:43
  • 1
    $\begingroup$ @FrankHarell I like Excel-bashing, R and literate programming as much as the next guy but spreadsheets help an awful lot of people actually getting stuff done. This should give pause to anybody wanting to push other tools. We don't even know if the OP is doing research at all (the data dictionary bit suggests that he might not be)… $\endgroup$ – Gala May 3 '13 at 20:59
  • $\begingroup$ I have to agree with @GaëlLaurans on this; some of us aren't in a position to need or justify learning R, but we still have need of a stats tool. Excel fills this need utilizing our existing skill set, software and data sets. $\endgroup$ – dav May 3 '13 at 22:25
  • 2
    $\begingroup$ @FrankHarell Maybe so but other objectives than reproducibility might be just as important including getting results ASAP, especially if you are not going to analyze data every week. Even the fact that more people know it is a valid reason to choose it over something else (or for employers to demand it). How “reproducible” is something implemented with an R script if you are the only one knowing R and you leave your job? Bottom line is that the mere fact that Excel is so popular shows it's providing something to its users. $\endgroup$ – Gala May 4 '13 at 6:34
  • 2
    $\begingroup$ The funny thing is that we only ever think about making such comments when it's about Excel. Nobody goes around writing “R is too slow”, “SAS is too expensive” or “Minitab does not run under Mac OS”, let alone “Math is too complicated for many people” under every question on this site. It's all just as true or at the very least defensible but it's not relevant or helpful. $\endgroup$ – Gala May 4 '13 at 6:36
4
$\begingroup$

Excel has a Microsoft provided Descriptive Statistics Tool in the Data Analysis Add-in. You'll need to activate it in your add-ins (varies based upon your version of Excel and you didn't mark which you use).

Here's a link loading the Analysis Toolpak for: Excel 2003, Excel 2007, or Excel 2010

In Excel 2010, once you've added it, you'll find Data Analysis in the Data Tab of the Ribbon. Among the tools they provide is Descriptive Statistics, which for any given range you provide will give you: Mean, Standard Error, Median, Mode, Standard Deviation, Sample Variance, Kurtosis, Skewness, Range, Minimum, Maximum, Sum, Count

There's also quite a few other tools for you if you prefer Excel to a more specialized tool like R.

$\endgroup$
3
$\begingroup$

Basic descriptive stats can be computed in Excel, either with functions like AVERAGE, MAX, MIN or using a table (formerly known as a list, see http://office.microsoft.com/en-001/excel-help/overview-of-excel-tables-HA010048546.aspx).

$\endgroup$
0
$\begingroup$

AnswerMiner is a good option as well. You can explore your data fast and easy. There is a Relation map and table in the app that shows you the relationships (how strong they are) in your data. Automatic charts are a huge help when it comes to visualize your data.

The prediction tree is also a good feature that builds decision trees automatically in a minute and helps you make prediction based on your past data.

You can try for free here.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.