isnan

欠損値はnp.nan

欠損値の判別はnp.isnanで，これを使って置き換えもできる．

#例

>>> x=np.array([1, 2, np.nan, 3, 4])

array([ 1., 2., nan, 3., 4.])

>>> np.isnan(x)

array([False, False, True, False, False], dtype=bool)

>>> x[np.isnan(x)]=0

>>> x

array([ 1., 2., 0., 3., 4.])

Missing value is np.nan

To find missing value, you can use np.isnan, and you can replace the missing value also using np.isnan.

#Example

>>> x=np.array([1, 2, np.nan, 3, 4])

array([ 1., 2., nan, 3., 4.])

>>> np.isnan(x)

array([False, False, True, False, False], dtype=bool)

>>> x[np.isnan(x)]=0

>>> x

array([ 1., 2., 0., 3., 4.])