| |
Methods defined here:
- __getslice__(self, i, j)
- !!! Do NOT create a new Matrice; so destruction of the
original Matrice entails destruction of this result
(x.__getslice__(i,j) <==> x[i:j]).
- __init__(self, **kw)
- Keyword argument:
fic -- Build from filename fic.
- backward(self, g, lx)
- Build of data g and Lexique lx, using backward algorithm,
with the logarithm of the computed probabilities.
- derivate(self, g)
- Build computing the differences (pos_i-pos_(i-1)) on data g.
- fb(self, g, lx)
- Build of data g and Lexique lx, using forward-backward
algorithm, with the logarithm of the computed probabilities.
- forward(self, g, lx)
- Build of data g and Lexique lx, using forward algorithm,
with the logarithm of the computed probabilities.
- generate(self, lg, l)
- Create an empty Matrice, with length lg and descriptor
names and numbers, chars, or #numbers, those of list l.
- integrate(self, g)
- Build computing the cumulated sums on data g, ie:
res[i]=\sum_{k=0}^i g[k].
- prediction(self, g, l)
- On Lexique l, build from data g. The descriptors are set
using the descriptor numbers of l, and the values at each position are
computed by the predictions of the corresponding descriptors.
The descriptor patterns are set such as the number of the line
corresponds with the descriptor used in first position.
- read_nf(self, nf)
- Build from filename nf.
- smoothe(self, g, sz=1, st=1, **kw)
- Build for data g, using sliding windows of size sz(=1),
with steps st(=1).
Keyword argument: f -- if equals 'm', compute the mean value,
otherwise compute the sum.
Methods inherited from _Matr:
- __add__(self, m)
- Return a Matrice = self + (Matrice m) if both Matrice fit
(x.__add__(m) <==> x+m).
- __iadd__(self, m)
- += (Matrice m) if Matrices fit (x.__iadd__(m) <==> x+=m).
- __imul__(self, x)
- *= (float x) (m.__imul__(x) <==> m*=x).
- __isub__(self, m)
- -= (Matrice m) if Matrices fit (x.__isub__(m) <==> x-=m).
- __len__(self)
- Return length of the Matrice (x.__len__() <==> len(x)).
- __str__(self)
- (x.__str__()<==> str(x))
- __sub__(self, m)
- Return a Matrice = self - (Matrice m) if Matrices fit
(x.__sub__(m) <==> x-m).
- copy(self)
- Build a NEW Matrice as a copy of the given one.
- desc(self)
- Return the list of descriptors.
- exp(self)
- Replace all the values by their exponential.
- g_val(self, v, c, i)
- Put value v at character c line i.
- line(self, i)
- Return a dictionnary {character: value of the character at
line i}.
- ln(self)
- Replace all the values by their neperian logarithm.
- max(self, i)
- Return the maximum value on descriptor i.
- ndesc(self)
- Return number of descriptors.
- set_proba(self)
- Normalize so that each line corresponds the logarithms of
a probability distribution.
- shuffle(self)
- Quasi-uniformly randomized using (len*(log(len)+1)/2)
transpositions.
- val(self, c, i)
- Return the value of character c at line i.
|