Overview
Function fields in the Chart Builder allow you to create a new field available to the chart, which is based on a report field with a selected function applied to it.
The more basic chart functions, such as Average, Trend, Forecast, and Accumulation can be applied to the advanced ones like Set Analysis or Variance.
Using Functions
When building a chart you can add a function one of two ways:
Location  Screenshot  Description 

Function List  You can drag a function from the function list on the left side of the chart builder screen. These fields will become available once required field components have been added to the chart. For example, some functions require a metric to be based on.  
Field List  You can add a function directly from the field you wish to base it on:

Selecting a Function
Once you have added a function to your chart, you can change the specific function being applied by clicking on the drop down menu on the field (hover the mouse over the field name).
Function Options
Once you have selected the specific function you wish to use, you can manually configure all of the parameters for it through the Series menu.
Average
Mean
This displays a line calculated as the sum of all values, divided by the number of values in a dataset.
Median
This displays a line calculated by ordering all the values in ascending order and using the middle value. This can be a more useful measure than mean if the dataset has extreme outliers.
Mode
This displays a line calculated by finding the value which occurs most in the dataset.
Trend
Auto Trend
Much like with auto charts, Yellowfin applies what it deems most useful on the particular chart based on some complex algorithms.
Option  Description 

Confidence Intervals  These are used to indicate a range where unknown or missing values are likely to fall. 
Interval Range (if confidence intervals used)  This specifies the certainty range of values to be displayed.

Moving Average
This displays the mean, calculated using values from a set numer of periods before each point. For example, if your dataset contains 10 periods, a moving average may be set to use 5 at a time. This can be used to follow changes in the data on a line hat is smoother than the actual values, making it useful when displaying trend for noisy (spikey) data.
Option  Description 

Periods  This specifies the number of periods the moving average will cover. A higher number of periods will result in a smoother, but less responsive trend line. 
Missing Values  This specifies how missing data should be treated.

Confidence Intervals  These are used to indicate a range where unknown or missing values are likely to fall. 
Interval Range (if confidence intervals used)  This specifies the certainty range of values to be displayed.

Linear Regression
This displays a straight line that indicates the relationship between the values on the x & y axes. This can be useful to give an idea of the general trend of data.
Option  Description 

Confidence Intervals  These are used to indicate a range where unknown or missing values are likely to fall. 
Interval Range (if confidence intervals used)  This specifies the certainty range of values to be displayed.

Polynomial Regression
This displays a curved line to the dataset which indicates the relationship between the values on the x & y axes. This can be used where the relationship between the values is not completely linear, for example if your trend has significant fluctuations through it.
Option  Description 

Order  This specifies how many terms will make up the polynomial for the functions. The higher the degree, the more tightly fitted the regression line. 
Confidence Intervals  These are used to indicate a range where unknown or missing values are likely to fall. 
Interval Range (if confidence intervals used)  This specifies the certainty range of values to be displayed.

Forecast
Auto Trend
Much like with auto charts, Yellowfin applies what it deems most useful on the particular chart based on some complex algorithms.
Option  Description 

Periods Forward  This specifies the number of periods into the future this function should forecast. The granularity of these periods is controlled by the overall granularity of the chart. 
Prediction Intervals  These are used to indicate a range where the actual value is likely to fall. 
Interval Range (if prediction intervals used)  This specifies the certainty range of values to be displayed.

Hide NonForecast Results  This specifies if forecast should be dispalyed either for the whole dataset range OR only after the end of the dataset range. 
Simple Exponential Smoothing
This displays a forecast based on the average of previous values in the dataset, with weighting defined by alpha (recent vs. historic data). This is useful or forecasting data which has no general trend or seasonality.
Option  Description 

Alpha  This specifies how much weight is given to recent data vs. older data. This higher the alpha value, the more weight is given to recent data. 
Periods Forward  This specifies the number of periods into the future this function should forecast. The granularity of these periods is controlled by the overall granularity of the chart. 
Missing Values  This specifies how missing data should be treated. 
Prediction Intervals  These are used to indicate a range where the actual value is likely to fall. 
Interval Range (if prediction intervals used)  This specifies the certainty range of values to be displayed.

Hide NonForecast Results  This specifies if forecast should be dispalyed either for the whole dataset range OR only after the end of the dataset range. 
Double Exponential Smoothing
This displays a forecast based on the average of previous values in the dataset, with weighting defined by alpha (recent vs. historic data) and beta (trend). This is useful for forecasting data which has a general trend, but no seasonality.
Option  Description 

Alpha  This specifies how much weight is given to recent data vs. older data. This higher the alpha value, the more weight is given to recent data. 
Beta  This specifies how much weight is given to the trend of the data. The higher the beta value, the more weight is given to the data's trend. 
Periods Forward  This specifies the number of periods into the future this function should forecast. The granularity of these periods is controlled by the overall granularity of the chart. 
Missing Values  This specifies how missing data should be treated. 
Prediction Intervals  These are used to indicate a range where the actual value is likely to fall. 
Interval Range (if prediction intervals used)  This specifies the certainty range of values to be displayed.

Hide NonForecast Results  This specifies if forecast should be dispalyed either for the whole dataset range OR only after the end of the dataset range. 
Triple Exponential Smoothing
This displays a forecast based on the average of previous values in the dataset, with weighting defined by alpha (recent vs. historic data), beta (trend), and gamma (seasonality). This is useful for forecasting data which has a general trend and varies seasonally. Triple exponential smoothing requires at least two years of data for its seasonality calculations.
Option  Description 

Alpha  This specifies how much weight is given to recent data vs. older data. This higher the alpha value, the more weight is given to recent data. 
Beta  This specifies how much weight is given to the trend of the data. The higher the beta value, the more weight is given to the data's trend. 
Gamma  This specifies how much weight is given to the seasonality (what happened in corresponding periods of previous years). The higher the gamma, the more weight is given to the data's seasonality. 
Periods Forward  This specifies the number of periods into the future this function should forecast. The granularity of these periods is controlled by the overall granularity of the chart. 
Missing Values  This specifies how missing data should be treated. 
Prediction Intervals  These are used to indicate a range where the actual value is likely to fall. 
Interval Range (if prediction intervals used)  This specifies the certainty range of values to be displayed.

Hide NonForecast Results  This specifies if forecast should be dispalyed either for the whole dataset range OR only after the end of the dataset range. 
Moving Average
This displays the mean, calculated using values from a set number of periods before each point. For example, if your dataset contains 10 periods, a moving average may be set to use 5 at a time. This can be used to follow changes in the data on a line that is smoother than the actual values, making it useful when displaying trend for noisy (spikey) data.
Option  Description 

Periods  This specifies the number of periods the moving average will cover. A higher number of periods will result in a smoother, but less responsive trend line. 
Periods Forward  This specifies the number of periods into the future this function should forecast. The granularity of these periods is controlled by the overall granularity of the chart. 
Missing Values  This specifies how missing data should be treated. 
Prediction Intervals  These are used to indicate a range where the actual value is likely to fall. 
Interval Range (if prediction intervals used)  This specifies the certainty range of values to be displayed.

Hide NonForecast Results  This specifies if forecast should be dispalyed either for the whole dataset range OR only after the end of the dataset range. 
Accumulation
This displays an accumulative % of total line on the chart, using a secondary axis.
Set Analysis
This allows you to display a subset of the data contained within a specified field.
In order to define a Set Analysis field you will:
 Select a report metric field to base the results on,
 Specify a range of filters that will be applied solely to this field within the chart
Option  Description 

Set Name  This specifies the display name of the resulting set analysis field, used in the chart labels and/or legend. 
Set Metric  This allows you to select the field to filter down to your specified set. Only metric fields included in your report will be available in this list. 
Filter  This allows you to specify the filters to be applied to your set field. Only category fields included in your report will be available in this list.

Note: Set Analysis filtering happens after the report results are returned from the database. This means that set analysis filters do not appear within the report SQL, and as such will be applied after user prompt filters.
Learn how to perform a Set Analysis function here.
Variance
This displays the difference between two fields from either the table or chart. The variance can be displayed as either the calculated value, or a % difference.
Option  Description 

Variance Name  This specifies the display name of the resulting variance field, used in the chart labels and/or legend. 
Variance Between  This allows you to select two fields or data sets to apply the variance calculation to. The values in the second field will be subtracted from the values in the first field, as part of the calculation, so order is important. These are the types of field available to be used as part of a variance (selected through the dropdown):
Or enable the Custom Set toggle to perform set analysis through the Create button. 
Display Variance as  This allows you to select how the resulting variance values are displayed. There are four options:

Learn how to use a Variance function here.