unittest::InflateUnitTestData - reformat compressed reference dataSynopsis:
array1 array2 InflateUnitTestData -> array3
% h= (in ms)
[ 0.1 0.2 0.5 1.0 ]
% time voltage
[ 1 -69.4229]
[ 2 1 -68.8515]
[ 3 -68.2858]
[ 4 2 -67.7258]
[ 5 1 -67.1713]
[ 6 3 -66.6223]
[ 7 -66.0788]
[ 8 4 -65.5407]
[ 9 -65.008]
[ 10 5 2 1 -64.4806]
[ 11 -63.9584]
The idea of the compressed data format is to represent the reference
data in a compact human readable format at the end of the testsuite
files. Although redundant, the time stamps of all step sizes
corresponding to a particular reference data point are listed.
array1 - array of sorted simulation step sizes [h1,...,hn]
All simulation step sizes must be multiples of the
smallest one h1.
array2 - array of reference data with one row per data point.
the last element v in each row is the value of the recorded
quantity. The first element in each row is the time stamp
t1 of the smallest step size. The time stamps of larger
step sizes are located between t1 and v if the temporal
position coincides with t1.
array3 - array of length n with one entry per step size h1,...,hn.
Each entry is a two-dimensional vector of tuples [ti,vi],
where ti is a time stamp and vi the corresponding value of
the recorded quantity.