| CPC G06F 30/20 (2020.01) [G06F 16/2453 (2019.01); G06F 16/248 (2019.01); G06F 16/283 (2019.01); G06F 2111/02 (2020.01)] | 20 Claims |

|
1. A method for modeling data, the method comprising:
storing data for a set of one or more models, wherein the data for each model of the set of one or more models is:
based on a multi-dimensional structure comprising a plurality of dimensions; and
stored, across the plurality of dimensions, as a one-dimensional vector;
accessing a particular subset of the data for the set of one or more models along a particular dimension of the multi-dimensional structure by:
computing an offset for one or more constraints of the particular dimension, wherein:
the one or more constraints is used to limit an accessed portion of the data; and
the offset moves a first element to the particular dimension that matches each constraint of the one or more constraints;
determining a step and a count for one or more constraint loops, wherein, for a given constraint loop of the one or more constraint loops:
the given constraint loop has a corresponding constraint of the one or more constraints;
the given constraint loop produces a set of indices for the corresponding constraint;
an associated count represents a quantity for the set of indices; and
an associated step represents a quantity of spaces between each of the set of indices; and
generating a non-linear arithmetic progression (FAP) series to access the particular subset of the data for the set of one or more models along the particular dimension of the multi-dimensional structure, wherein the non-linear FAP series is based, at least in part, on the one or more constraint loops and the offset;
performing calculations for the multi-dimensional structure of at least one model of the set of one or more models based on the particular subset of the data; and
providing results of the calculations to a client application configured to display the results.
|