On the maximum-entropy estimate of the electron density function

The principle of maximum entropy, considered as a form of statistical inference, is used to obtain an estimate of the electron density function on the basis of partial information. First a maximum-entropy probability distribution of maps, which explicitly takes into account the available information, is obtained, its functional form being a strict consequence of the type of constraint used. Next the electron density function is estimated using this probability distribution. For the particular type of constraint considered here the formulation presented is shown to correspond exactly to a maximum-entropy algorithm using a new form of the configurational entropy of maps.