### Materials and Methods

The chile peppers that the program will be based on are the Bell Pepper, Piquin, Jalapeno, and Sandia.  To create the computer program we have gathered data from all types of sources.  Most of these sources were online articles written by scientists or students who are interested in chile peppers.  One of the most vital sources of information that we collected was an article published by the Chile Pepper Institute titled Yield and Quality of Trickle Irrigated Chile.  This article described an experiment done by P.J. Wierenga that studies how water affects the chile’s yield and its pungency.

In order to be able to work with the data, we had to convert the measurements into Scoville Heat Units.  Most of the studies were all measured in g/mg of capsaicin.  In order to convert all the measurements we talked to Dr. Bosland to determine how to convert g/mg of capsaicin into SHU, Scoville Heat Units.  To convert mg/g to SHU, you need to know that 1 mg/1g is equal to 1000 parts per million (ppm).  You need to multiply the ppm by 16 to convert everything into SHU.

To covert mg/g of capsaicin into SHU use this formula

 1g / 1mg = 1000 ppm   16 x  ppm = SHU value

After we had all of the data converted, we began to create graphs using Microsoft Excel. We analyzed the graphs to determine if there was any correlation between any of the factors.  We noticed that the pungency of the chile is inversely proportional with the yield.  If the chile peppers were hot the plant did not give as much fruit as in a mild chile.  The amount of water given and the pungency of peppers are also inversely proportional.  The more water was given the less pungent the chile is.

When all of the graphs were completed, we added a best fit curve to be able to get the equation of the line and the R2 value.  We did this because we did not have much data that compared the amount of water to the pungency of the chile.  The R2 value is the probability of having the next data point on the line.  With the R2 value we could determine where the next set of data points would fall on the line.  The closer the R 2 value is to 1, the higher the probability of having the next data point on the line.

We also used the equations and the R2 values generated from the Excel graphs, to add randomness to the program.  This was done by choosing the equations that graphed pungency against amount of water added, and had an R2 value closest to 1.  We added all of these equations to the program and then let the program decide which equation it wants to use, to determine the pungency of the plant being modeled.  We decided to do this because we noticed that many other programs just used ± 10% as randomness.  We wanted to try a newer method of randomization so we picked this method.  This way the results will never be the same twice.

After we had all of the data and had completed the program, we decided to include more variables into the program.  The variables are height of plant, position of flower, and fruit length.  This data came from Dr. Bosland and the article by P.J. Wierenga.  We had to do some extra research on how the different chile types grow.  It did not take us long to come across a study that explained that most chile varieties all grow the same.  This came to our aid because due to this we did not need to add any type of randomness to this part of the program.