Collecting data and livin'

Another day another dollarrrr. What do we have here. Ok. Running in to a lot of problems in which the customer is making changes to the capability of the product. Functions were written to apply to certain areas, and the customer is pushing the boundaries of these areas. This is creating several problems because there were predefined fields that aren't dynamic. Beyond the problem of fixing the problems he creates, the biggest problem is THE CUSTOMER's EXPECTATIONS I was put on this project a few weeks ago to help with the code because no one else in the group knows VBA except me and a guy who just left to go on maternity leave. It means I have a somewhat superficial understanding of how the product is organized, which means I am susceptible to making fixes to cosmetic problems that end up creating more problems. After a few weeks, I'm much more careful with the types of fixes I make. I've learned from my homeslice Joe to minimize the customer's ability...

1) Woops, someone made me realize I have some "ordinal" categorical variables: Size: vlow,low,med,large,vlarge. There is an order to their naming. vlow is furthest from vlarge vlarge is closer to large etc. The question is, how to convert them to numeric values with proportional distance relationships. PRINCIPLES "Statistical moments": Shape of a set of points Ex. Points representing probability density 0th moment: total probability ie. 1 1st moment: mean 2nd moment: variance 3rd moment: skewness 4th moment (with normalization and shift): kurtosis (high or low arch)

I'm just really living a full life now, and wanted everyone to know what that means, today.   1) I wanted to graph a custom line on an excel scatter plot. I added a new column on my raw data sheet and for every x value computed the y using my regression equation. Then, I graphed the points and chose the "Show trend line" option. It's not perfect since I'd rather it just show the line and not the made up points underneath, but it works.   2) I want to track down how the data miner we hired filled in missing data. IT WAS TOO COMPLICATED AND NOT WELL DOCUMENTED. Off to the interwebs. I've been using Weka lately. Weka is a collection of machine learning algorithms for data mining tasks. Weka contains tools for data pre-processing, classification, regression, clustering, association rules, and visualization. Weka is a New Zealand bird with an inquisitive nature. I like the word inquisitive. It's like smooooooooth curiosity. 3) Figuring out which...