View source for ArgMin and ArgMax

[[Category:Array-reducing functions]]
[[Category:Doc Status C]] <!-- For Lumina use, do not change -->

= Functions ArgMax(x, i) and ArgMin(x, i)=

ArgMax(x, i) returns the value of index i corresponding to the largest value from array x, which must be indexed by i. More specifically, it returns a value, ix where x[i =ix] = Max(x, i). If there are ties, i.e. x has two or more values that are equal largest, it returns the value nearest the end of i.  Similarly, ArgMin(x, i) returns the value of i for which x is ''smallest''. 

ArgMin is new to Analytica 4.0.  

== Declaration ==

 ArgMin( R : vector[I] ; I : ... optional IndexType ; position : optional boolean )
 ArgMax( R : vector[I] ; I : ... optional IndexType ; position : optional boolean )

== When x has multiple indexes ==

If array x has an index or indexes in addition to i, Argmax(x, i) or Argmin(x, i) return an array indexed by those other indexes, but not by i.  For example, if x is indexed by i and j, it will return the value of i for which x is largest (or smallest) for each value of j.

== Searching over Multiple Indexes ==

(new to 4.0)

If you want to find the smallest or largest value of x over more than one index, you can list all those indexes: e.g.
 ArgMax(X, I: Mpg, Car_type)

It will then return an array dimensioned by a local index .Dim and any indexes of X that you did not list as parameters. Index .Dim will contain the names of the indexes given as parameters, Mpg and Car_type, in this case. The result will contain the values of Index Mpg and Car_type for which X is maximum (or minimum).

== Positional Usage ==

By default, ArgMax (ArgMin) returns a value of the index.  If you specify optional parameter position as true, they return the position of the value in the index for which x is maximum (minumum) -- a number.  For example:
  ArgMax(R: Profit, I: UnitPrice, Position: true)
This helps avoid ambiguity when the index contains duplicate values.

== Related Functions ==

* [[Min]] and [[Max]]
* [[CondMin]] and [[CondMax]]
* [[SubIndex]] and [[PositionInIndex]]