GitLab

Following a discussion with J. V. Heiden.

See #3590.

See #3572 and #3573.

See #2828 and #3457.

See #3358

See #3575.

See #3358.

Extracting getClone() to SampleOutput.

See #3358.

See #3358.

See #3358.

See #3358.

See #3358.

See #3358.

See #3358.
Far from perfect, but allows to gather code.

There is a problem in g++4.8: erase with a const_iterator doesn't exist
(while it should: it is c++11).

A higher p-value threshold will give a lower limit for number of k-mers hence will be more sensitive (but less time-efficient).
See #3225

This will ease unit testing

Rather we use an uncertainty range (see #3225 for why and how).

It is necessary to get the length of the longest transferred sequence as it
will be used as a reference to compute the empirical probablity that a k-mer
has matched in the longest possible sequence. Therefore we make the assumption
that k-mers matched in the longest sequence but we have no way of knowing as
we don't keep track of where the k-mers matched.

Fixes #3225

Compute the minimal number of k-mers of an uncertainty range whose mean is
given by the empirical number of k-mers found.

The probability is provided as an integer as we don't have Z-scores for every
possible value and float comparison can be a bit tricky.

Factor a code that what called 4 times.

Fixes #3395

See #3393

Will be called when not all the reads can be stored.

See #3393

This will help to reuse those functions elsewhere in the code

See #3393

For more informations, see #1409.