Interpoluj wartości NaN w tablicy numpy

Czy istnieje szybki sposób zastąpienia wszystkich wartości NaN w tablicy numpy (powiedzmy) wartościami interpolowanymi liniowo?

Na przykład,

[1 1 1 nan nan 2 2 nan 0]

Zostanie zamienione na

[1 1 1 1.3 1.6 2 2  1  0]
Author: ppperry, 2011-06-29

8 answers

Zdefiniujmy najpierw prostą funkcję pomocniczą, aby ułatwić obsługę indeksów i indeksów logicznych NaNs:

import numpy as np

def nan_helper(y):
    """Helper to handle indices and logical indices of NaNs.

    Input:
        - y, 1d numpy array with possible NaNs
    Output:
        - nans, logical indices of NaNs
        - index, a function, with signature indices= index(logical_indices),
          to convert logical indices of NaNs to 'equivalent' indices
    Example:
        >>> # linear interpolation of NaNs
        >>> nans, x= nan_helper(y)
        >>> y[nans]= np.interp(x(nans), x(~nans), y[~nans])
    """

    return np.isnan(y), lambda z: z.nonzero()[0]

Teraz nan_helper(.) można teraz używać w następujący sposób:

>>> y= array([1, 1, 1, NaN, NaN, 2, 2, NaN, 0])
>>>
>>> nans, x= nan_helper(y)
>>> y[nans]= np.interp(x(nans), x(~nans), y[~nans])
>>>
>>> print y.round(2)
[ 1.    1.    1.    1.33  1.67  2.    2.    1.    0.  ]

---
Chociaż może się wydawać, że najpierw trochę przesadziłem, aby określić oddzielną funkcję do robienia takich rzeczy:

>>> nans, x= np.isnan(y), lambda z: z.nonzero()[0]
W końcu wypłaci dywidendę.

Więc, ilekroć pracujesz z danymi związanymi z Nan, po prostu hermetyzuj wszystkie (new Nan related) potrzebna funkcjonalność, pod pewnymi konkretnymi funkcjami pomocniczymi. Twoja baza kodu będzie bardziej spójna i czytelna, ponieważ podąża za łatwo zrozumiałymi idiomami.

Interpolacja rzeczywiście jest miłym kontekstem, aby zobaczyć, jak odbywa się obsługa NaN, ale podobne techniki są wykorzystywane również w różnych innych kontekstach.
 65
Author: eat,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2016-04-22 17:14:54

Wymyśliłem ten kod:

import numpy as np
nan = np.nan

A = np.array([1, nan, nan, 2, 2, nan, 0])

ok = -np.isnan(A)
xp = ok.ravel().nonzero()[0]
fp = A[-np.isnan(A)]
x  = np.isnan(A).ravel().nonzero()[0]

A[np.isnan(A)] = np.interp(x, xp, fp)

print A

Drukuje

 [ 1.          1.33333333  1.66666667  2.          2.          1.          0.        ]
 19
Author: Petter,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2011-06-29 10:19:56

Wystarczy użyć numpy logical and there where, aby zastosować interpolację 1D.

import numpy as np
from scipy import interpolate

def fill_nan(A):
    '''
    interpolate to fill nan values
    '''
    inds = np.arange(A.shape[0])
    good = np.where(np.isfinite(A))
    f = interpolate.interp1d(inds[good], A[good],bounds_error=False)
    B = np.where(np.isfinite(A),A,f(inds))
    return B
 8
Author: BRYAN WOODS,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2017-05-17 19:07:56

Może być łatwiej zmienić sposób generowania danych, ale jeśli nie:

bad_indexes = np.isnan(data)

Tworzenie tablicy logicznej wskazującej gdzie są Nan

good_indexes = np.logical_not(bad_indexes)

Utwórz tablicę logiczną wskazującą, gdzie znajduje się obszar dobrych wartości

good_data = data[good_indexes]

Ograniczona wersja oryginalnych danych z wyłączeniem Nan

interpolated = np.interp(bad_indexes.nonzero(), good_indexes.nonzero(), good_data)

Uruchom wszystkie złe indeksy przez interpolację

data[bad_indexes] = interpolated

Zastąp oryginalne dane wartościami interpolowanymi.

 4
Author: Winston Ewert,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2011-06-29 10:10:58

Lub bazując na odpowiedzi Winstona

def pad(data):
    bad_indexes = np.isnan(data)
    good_indexes = np.logical_not(bad_indexes)
    good_data = data[good_indexes]
    interpolated = np.interp(bad_indexes.nonzero()[0], good_indexes.nonzero()[0], good_data)
    data[bad_indexes] = interpolated
    return data

A = np.array([[1, 20, 300],
              [nan, nan, nan],
              [3, 40, 500]])

A = np.apply_along_axis(pad, 0, A)
print A

Wynik

[[   1.   20.  300.]
 [   2.   30.  400.]
 [   3.   40.  500.]]
 3
Author: user423805,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2012-08-22 12:52:46

Dla danych dwuwymiarowych, SciPy griddata działa dość dobrze dla mnie:

>>> import numpy as np
>>> from scipy.interpolate import griddata
>>>
>>> # SETUP
>>> a = np.arange(25).reshape((5, 5)).astype(float)
>>> a
array([[  0.,   1.,   2.,   3.,   4.],
       [  5.,   6.,   7.,   8.,   9.],
       [ 10.,  11.,  12.,  13.,  14.],
       [ 15.,  16.,  17.,  18.,  19.],
       [ 20.,  21.,  22.,  23.,  24.]])
>>> a[np.random.randint(2, size=(5, 5)).astype(bool)] = np.NaN
>>> a
array([[ nan,  nan,  nan,   3.,   4.],
       [ nan,   6.,   7.,  nan,  nan],
       [ 10.,  nan,  nan,  13.,  nan],
       [ 15.,  16.,  17.,  nan,  19.],
       [ nan,  nan,  22.,  23.,  nan]])
>>>
>>> # THE INTERPOLATION
>>> x, y = np.indices(a.shape)
>>> interp = np.array(a)
>>> interp[np.isnan(interp)] = griddata(
...     (x[~np.isnan(a)], y[~np.isnan(a)]), # points we know
...     a[~np.isnan(a)],                    # values we know
...     (x[np.isnan(a)], y[np.isnan(a)]))   # points to interpolate
>>> interp
array([[ nan,  nan,  nan,   3.,   4.],
       [ nan,   6.,   7.,   8.,   9.],
       [ 10.,  11.,  12.,  13.,  14.],
       [ 15.,  16.,  17.,  18.,  19.],
       [ nan,  nan,  22.,  23.,  nan]])

Używam go na obrazach 3D, operując na plasterkach 2D (4000 plasterków 350x350). Cała operacja trwa jeszcze około godziny : /

 2
Author: Gilly,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2016-06-17 13:23:40

Bazując na odpowiedzi autorstwa Bryana Woodsa , zmodyfikowałem jego kod, aby również konwertować listy składające się tylko z NaN na listę zer:

def fill_nan(A):
    '''
    interpolate to fill nan values
    '''
    inds = np.arange(A.shape[0])
    good = np.where(np.isfinite(A))
    if len(good[0]) == 0:
        return np.nan_to_num(A)
    f = interp1d(inds[good], A[good], bounds_error=False)
    B = np.where(np.isfinite(A), A, f(inds))
    return B

Prosty dodatek, mam nadzieję, że komuś się przyda.

 2
Author: rbnvrw,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2017-05-23 12:02:53

Potrzebowałem podejścia, które również wypełniłoby NaN na początku końca danych, czego główna odpowiedź nie wydaje się robić.

Funkcja, którą wymyśliłem, wykorzystuje regresję liniową do wypełnienia NaN. to rozwiązuje mój problem:

import numpy as np

def linearly_interpolate_nans(y):
    # Fit a linear regression to the non-nan y values

    # Create X matrix for linreg with an intercept and an index
    X = np.vstack((np.ones(len(y)), np.arange(len(y))))

    # Get the non-NaN values of X and y
    X_fit = X[:, ~np.isnan(y)]
    y_fit = y[~np.isnan(y)].reshape(-1, 1)

    # Estimate the coefficients of the linear regression
    beta = np.linalg.lstsq(X_fit.T, y_fit)[0]

    # Fill in all the nan values using the predicted coefficients
    y.flat[np.isnan(y)] = np.dot(X[:, np.isnan(y)].T, beta)
    return y

Oto przykładowy przypadek użycia:

# Make an array according to some linear function
y = np.arange(12) * 1.5 + 10.

# First and last value are NaN
y[0] = np.nan
y[-1] = np.nan

# 30% of other values are NaN
for i in range(len(y)):
    if np.random.rand() > 0.7:
        y[i] = np.nan

# NaN's are filled in!
print y
print linearly_interpolate_nans(y)
 1
Author: nlml,
Warning: date(): Invalid date.timezone value 'Europe/Kyiv', we selected the timezone 'UTC' for now. in /var/www/agent_stack/data/www/doraprojects.net/template/agent.layouts/content.php on line 54
2016-08-30 09:23:01