dataset.value has been deprecated since h5py version 1.3.

From http://mail.scipy.org/pipermail/scipy-user/2010-February/024364.html

[SciPy-User] [ANN] HDF5 for Python (h5py) 1.3.0 beta
Andrew Collette andrew.collette@gmail....
Tue Feb 23 15:52:12 CST 2010

...
 - Dataset .value attribute is deprecated.  Use dataset[...] or dataset[()].
...

diff --git a/posts/HDF5.mdwn b/posts/HDF5.mdwn
index f3760e69806f5a632674dd852be7e0a55246df3e..ff50d91f7195771f6ffdc13f4fba43f4ba6f9a23 100644 (file)
--- a/posts/HDF5.mdwn
+++ b/posts/HDF5.mdwn
@@ -80,7 +80,7 @@ Reading from and writing to a dataset
     >>> dset = f['dset']
     >>> dset
     <HDF5 dataset "dset": shape (4, 6), type "<i4">
-    >>> dset.value
+    >>> dset[...]
     array([[ 0,  1,  2,  3,  4,  5],
            [ 6,  7,  8,  9, 10, 11],
            [12, 13, 14, 15, 16, 17],
@@ -239,10 +239,10 @@ Just use the [Numpy slice indexing][slice] you're used to.
 
     >>> import h5py
     >>> import numpy
-    >>> f = h5py.File('hype.h5', 'w')
+    >>> f = h5py.File('slice.h5', 'w')
     >>> f['IntArray'] = numpy.ones((8, 10))
     >>> dset = f['IntArray']
-    >>> dset.value
+    >>> dset[...]
     array([[ 1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.],
            [ 1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.],
            [ 1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.],
@@ -252,7 +252,7 @@ Just use the [Numpy slice indexing][slice] you're used to.
            [ 1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.],
            [ 1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.,  1.]])
     >>> f['IntArray'][:,5:] = 2
-    >>> dset.value
+    >>> dset[...]
     array([[ 1.,  1.,  1.,  1.,  1.,  2.,  2.,  2.,  2.,  2.],
            [ 1.,  1.,  1.,  1.,  1.,  2.,  2.,  2.,  2.,  2.],
            [ 1.,  1.,  1.,  1.,  1.,  2.,  2.,  2.,  2.,  2.],
@@ -262,7 +262,7 @@ Just use the [Numpy slice indexing][slice] you're used to.
            [ 1.,  1.,  1.,  1.,  1.,  2.,  2.,  2.,  2.,  2.],
            [ 1.,  1.,  1.,  1.,  1.,  2.,  2.,  2.,  2.,  2.]])
     >>> dset[1:4,2:6] = 5
-    >>> f['IntArray'].value
+    >>> f['IntArray'][...]
     array([[ 1.,  1.,  1.,  1.,  1.,  2.,  2.,  2.,  2.,  2.],
            [ 1.,  1.,  5.,  5.,  5.,  5.,  2.,  2.,  2.,  2.],
            [ 1.,  1.,  5.,  5.,  5.,  5.,  2.,  2.,  2.,  2.],
@@ -273,7 +273,27 @@ Just use the [Numpy slice indexing][slice] you're used to.
            [ 1.,  1.,  1.,  1.,  1.,  2.,  2.,  2.,  2.,  2.]])
     >>> f.close()
 
+Here's an example of altering a scalar value:
+
+    >>> import h5py
+    >>> import numpy
+    >>> f = h5py.File('scalar.h5', 'w')
+    >>> f['int'] = 1
+    >>> dset = f['int']
+               >>> f['int'][...]
+    1
+               >>> f['int'][...] = 2
+               >>> f['int'][...]
+               2
+               >>> f.pop('int')
+    >>> f.close()
+
+I haven't been able to track down official documentation for the
+`dataset[...]` syntax, but it is mentioned in [the 1.3 release
+announcement][message] that Andrew sent to the `scipy-user` list.
+
 [slice]: http://docs.scipy.org/doc/numpy/reference/arrays.indexing.html
+[message]: http://mail.scipy.org/pipermail/scipy-user/2010-February/024364.html
 
 Datatypes
 ---------
@@ -285,12 +305,12 @@ Your array's `numpy.dtype` will be preserved.
     >>> f['complex'] = 2 + 3j
     >>> f['complex'].dtype
     dtype('complex128')
-    >>> type(f['complex'].value)
+    >>> type(f['complex'][...])
     <type 'complex'>
     >>> f['complex array'] = [1 + 2j, 3 + 4j]
     >>> f['complex array'].dtype
     dtype('complex128')
-    >>> type(f['complex array'].value)
+    >>> type(f['complex array'][...])
     <type 'numpy.ndarray'>
     >>> f.close()
 
@@ -358,7 +378,7 @@ Extendible datasets must be chunked.
     (2,)
     >>> c[:] = [9, 8, 7]
     >>> c.resize((6,))
-    >>> c.value
+    >>> c[...]
     array([1, 2, 3, 0, 0, 0])
     >>> c.resize((6,2))
     Traceback (most recent call last):