Percent Difference from Median [%Diff_Med]

Return to Complete List of Functions

The Percent Difference from Median function calculates the percent difference between a value and its median value. Parameters ------------------ Data          The data to analyze. This is typically a field in a data series or a calculated value. Period        The number of bars of data to include in the median, including the current value.                   For example, a period of 3 includes the current value and the two previous values. Function Value ------------------------ This function determines the median value of a data series within a given period. The median is the number in the middle of all of the values if the values were to be sorted. If an odd value is used for the period, the median will be an actual number from the series. If an even value is used, the median will be the average of the two middle values. If the data series does not have enough data to fill the period or it contains a null during the period, a null is returned. As a result of finding the median, half of the values during the given period have a value equal to or greater than the median and half of the values have a value equal to or less than the median. The median is useful for obtaining an "average" value of a data series while practically removing the effects of outliers. The percent difference from the median is the median subtracted from the current value divided by the median. Since this function uses division, if the median is 0, a null value will be returned. Usage ----------- Medians and moving averages are useful for smoothing noisy raw data, such as daily prices. Price data can vary greatly from day-to-day, obscuring whether the price is going up or down over time. By looking at the median of the price, a more general picture of the underlying trends can be seen. Since medians can be used to see trends, they can also be used to see whether data is bucking the trend. This makes the percent difference from the median useful for highlighting where the data is breaking away from the trend.