We begin our explorative analysis by calculating the maximum, mean, and minimum of the newly imported table. MATLAB calculates these statistics independently for each column in the table. For this purpose, we use three useful functions: max(), mean(),and min().

To find the maximum value of oxide content (from the third column to the eighth, representing oxide content of Na, Mg, Al, Si, K, and Ca), using curly braces, {}, simply type the following command:

>> Max = max(GlassIdentificationDataSet{:,3:8})
Max =
17.3800    4.4900    3.5000   75.4100    6.2100   16.1900   

This way, we get the maximum values in each of the specified columns. To calculate the mean of the same columns, we will write the command shown here:

>> Mean = mean(GlassIdentificationDataSet{:,3:8})
Mean =
13.4079    2.6845    1.4449   72.6509    0.4971    8.9570   

To find the minimum value in each column, we use the following command:

>> Min = min(GlassIdentificationDataSet{:,3:8})
Min =
10.7300         0    0.2900   69.8100         0    5.4300        

It may be useful to identify the records in which minimum and maximum are found; to get this information, just specify a second output parameter to return the row index. For example:

>> [Max,IndRowMax] = max(GlassIdentificationDataSet{:,3:8})
Max =
17.3800    4.4900    3.5000   75.4100    6.2100   16.1900   
IndRowMax =
185     1   164   185   172   108  

>> [Min,IndRowMin] = min(GlassIdentificationDataSet{:,3:8})
Min =
10.7300         0    0.2900   69.8100         0    5.4300 
IndRowMin =
107   106    22   107    64   186    

Additionally, using MATLAB's built-in functions, we can calculate median and mode. Recall that the median represents the intermediate value between the extremes of a set of data. MATLAB calculates the median for each column with the median() function:

>> Median = median(GlassIdentificationDataSet{:,3:8})
Median =
13.3000    3.4800    1.3600   72.7900    0.5550    8.6000        

The mode is the value that appears most often in a set of data. MATLAB calculates the mode for each column with the mode() function:

>> Mode = mode(GlassIdentificationDataSet{:,3:8})
Mode =
13.0000         0    1.5400   72.8600         0    8.0300        

When there are multiple values occurring equally frequently, mode returns the smallest of those values.