Ion Alberdi

doc: exact fhmm ? fhmm vs additive fhmm vs additive fhmm for nilm

6

Hello to all, when reading the source code of the exact_fhmm, it seems it's based on an **additive fhmm** whose hidden states are **decoded** using viterbi's algorithm. When reading related...

Outputs of the Direct Grammar in the README are misleading

2

Executing the Direct Grammar on a worksheet ```scala import parseback._ import parseback.compat.cats._ import cats.Eval implicit val W = Whitespace("""\s+"""r) lazy val expr: Parser[Int] = ( expr ~ "+" ~ term...

joblib not found when Parallel is executed inside a pex

8

Trying to package joblib inside a pex, executing the script ```bash #!/bin/bash python3.6 -m venv . . bin/activate # gets pip 19.0.3 pip install --upgrade pip pip install pex pex...

Use seed in train_test_split

1

Permits having the same result when running the same experiment twice. Follows recommendation given at https://github.com/coreylynch/pyFM/issues/19

Use https://cython.readthedocs.io/en/latest/src/userguide/numpy_tutorial.html#efficient-indexing-with-memoryviews instead of casting the arrays.data attribute into a DOUBLE*. The former strategy generated segfaults in some servers.

Make loky backend available in a pex

2

Fix #849