\(\newcommand{\W}[1]{ \; #1 \; }\) \(\newcommand{\R}[1]{ {\rm #1} }\) \(\newcommand{\B}[1]{ {\bf #1} }\) \(\newcommand{\D}[2]{ \frac{\partial #1}{\partial #2} }\) \(\newcommand{\DD}[3]{ \frac{\partial^2 #1}{\partial #2 \partial #3} }\) \(\newcommand{\Dpow}[2]{ \frac{\partial^{#1}}{\partial {#2}^{#1}} }\) \(\newcommand{\dpow}[2]{ \frac{ {\rm d}^{#1}}{{\rm d}\, {#2}^{#1}} }\)

base_hash

View page source

Base Type Requirements for Hash Coding Values

Syntax

code = hash_code ( x )

Purpose

CppAD uses a table of Base type values when recording AD < Base > operations. A hashing function is used to reduce number of values stored in this table; for example, it is not necessary to store the value 3.0 every time it is used as a Parameter .

Default

The default hashing function works with the set of bits that correspond to a Base value. In most cases this works well, but in some cases it does not. For example, in the base_adolc.hpp case, an adouble value can have fields that are not initialized and valgrind reported an error when these are used to form the hash code.

x

This argument has prototype

const Base & x

It is the value we are forming a hash code for.

code

The return value code has prototype

unsigned short code

It is the hash code corresponding to x . This intention is the commonly used values will have different hash codes. The hash code must satisfy

code < CPPAD_HASH_TABLE_SIZE

so that it is a valid index into the hash code table.

inline

If you define this function, it should declare it to be inline , so that you do not get multiple definitions from different compilation units.

Example

See the base_alloc hash_code and the adouble hash_code .