Generalisation of Levelt's Rank Reduction Algorithm 

Our generalizations of Levelt's algorithm (1990) compute an equivalent system to that of the input matrix, which has the lowest possible rank of singularity in the parameter epsilon (sigma=0) and algebraic rank for the leading coefficient w.r.t. this parameter. The output is the equivalent system, the transformation, and computation time.
Function calls: 

LeveltParam(Input matrix with the poles in the parameter, the variable, the parameter, truncation in the variable (in case of power series entries), truncation in parameter).  

LeveltParamM(Input matrix with the poles in the parameter, the variable, the parameter, truncation in the variable (in case of power series entries), truncation in parameter). LeveltParamM uses Moser's reduction criterion to avoid unnecessary computations 

>

 

(4.1)

 

>

 

(4.2)

 

>