There are many probability distributions developed by statisticians to model random data of different kinds, ranging from business data, finance data (stock prices) to engineering data (system failures) and environmental data (max. flood flows). While the standard Excel package includes some basic statistical functions, its support for probability distributions is very limited and almost useless for real world modelling applications. This article discusses the worksheet functions provided by EasyFitXL, the distribution fitting add-in for Excel that can be applied to perform a range of decision-making calculations using a variety of probability distributions… read the full article
In EasyFit 3.0 – back in 2006 – we introduced StatAssist, the built-in distribution viewer tool that closely integrates with the distribution fitting features of EasyFit. Since then, StatAssist has proven to be quite a useful feature, so we decided to include it into EasyFitXL, our distribution fitting add-in for Excel.
StatAssist displays graphs and other useful properties of all the probability distributions available in EasyFit. Even though it has initially been designed as an essential part of EasyFit, StatAssist can be used as a separate application – for example, to take a quick look at a distribution curve, or to calculate the distribution statistics… read the full article
It’s been a while since we updated our blog – that’s because we have been working on a new version of EasyFit planned to be released later in October 2008. Below is a quick summary of new features in EasyFit.
(Note: The price for the new version is likely to increase, but since all licenses come with one year of free updates, you can still order EasyFit at the current price this month and upgrade to the upcoming new version for free.)
New Probability Distributions
EasyFit will include 15 new distributions:
The Burr (Singh-Maddala), Dagum (Inverse Burr), Pareto Type II (Lomax), and Log-Gamma distributions are mainly used in economics and actuarial science. We have had numerous requests for these models from some of our customers – actuaries and finance industry professionals.
The Generalized Gamma distribution is an advanced 4-parameter model including the Gamma, Weibull, and Half-Normal distributions as special cases.
The Pearson Type V (Inverse Gamma) and Pearson Type VI (Beta Prime) distributions are a part of the Pearson family of distributions often used in financial applications.
The Levy distribution is one of the few distributions that are stable and that have probability density functions that are analytically expressible, the others being the Normal distribution and the Cauchy distribution.
The Log-Pearson III (often referred to as “LP3”) distribution is the standard model recommended by the U.S. Water Resources Council for use by all U.S. Government agencies for flood frequency analysis.
The Nakagami (Nakagami-m) and Rice (Nakagami-n) distributions are used in communications theory to model signal fading under specific conditions.
The Kumaraswamy distribution is used in electrical, civil, mechanical, and financial engineering applications dealing with variables that are lower and upper bounded.
Some other new models include the Hyperbolic Secant, Error (Exponential Power), and Reciprocal distributions.
New Excel Functions
In EasyFit 4.0, we added the ability to analyze probability data in Excel, which has proven to be quite a useful feature. EasyFit 5.0 (Professional Edition) extends this functionality by introducing several new functions making it even easier to use distributions in your worksheet models.
The DistFit function allows to fit a specified distribution to your data and display the resulting model right in a worksheet. This function can be useful if you already know the underlying distribution (e.g. the Normal model) and only need to estimate the distribution parameters. The return value of this function can be supplied to several other new functions such as DistPdf, DistCdf, DistRand etc. For example, specifying
=DistCdf(DistFit(“Normal”; B1:B100); 2)
in a worksheet will fit the Normal distribution to the data in cells B1:B100, and then evaluate the CDF of the resulting Normal distribution at x=2.
Click here to watch the quick Flash demo showing how to fit probability distributions using EasyFit and apply the best fitting distribution to perform specific calculations – for instance, make estimates using the quantile function, and calculate probabilities.
The data set used in this demo consists of maximum daily wind gust speeds recorded at Station TPLM2 located in the Atlantic Ocean during 2005-2007. This station is owned and maintained by the National Data Buoy Center, and measures wind speed, air temperature, sea temperature, and other data used for weather forecasting.
NOAA defines a wind gust as “a sudden, brief increase in speed of the wind” which usually lasts for less than 20 seconds. The relatively rare but very high wind gusts cause the most damage, that is why they are of more interest than the average daily wind speeds. In essence, the wind gusts are extreme events – no wonder the Generalized Extreme Value distribution perfectly fits the data:
(the x axis units are m/s)