`rec-array` conformance between anndataR and anndata(py)

[Artur-man]

There is a difference between rhdf5, Rarr and h5py on how rec arrays and structured data types are read.

To me, h5py and Rarr reads it correctly, but rhdf5 does not:
Huber-group-EMBL/rhdf5#192

Originally posted by @Artur-man in #190 (comment)

Will produce an example below now. However, it looks like the issue rises mostly from rhdf5 and h5py.

[Artur-man]

> adata$uns$rank_genes_groups$logfoldchanges
$`0`
  [1]  1.209112644  1.001628280  1.251380801  0.866736710  0.873581290  0.835230410  0.823304236  0.731522799  0.632428885  0.791670740
 [11]  0.483177572  0.569251180  0.533949852  0.510987341  0.396106869  0.404166162  0.423241198  0.372831702  0.395179331  0.325538129
 [21]  0.361550868  0.397086114  0.316921383  0.219644874  0.241885915  0.202931836  0.305081934  0.276373237  0.205611467  0.201653823
 [31]  0.168201476  0.203728065  0.183869943  0.156049907  0.141745120  0.139942512  0.093910255  0.103688799  0.132829919  0.129819363
...
$`5`
  [1]  0.83798558  1.26184928  0.60688692  1.34007716  0.81900942  0.58131063  0.80614799  0.53503847  0.88951248  1.04317546  0.46853259
 [12]  0.84463811  0.78316927  0.87504202  0.47333276  0.70802909  0.77588779  0.60755169  0.58391559  0.59059560  0.72878522  0.62302238
 [23]  0.47069362  0.49771789  0.64095032  0.41865152  0.39888418  0.52392238  0.65389711  0.33132794  0.58317381  0.34273294  0.32106584
 [34]  0.23855770  0.42743140  0.34370065  0.28144893  0.11881924  0.23293939  0.15941593  0.15797473  0.20382275  0.10664603  0.08825362
 [45]  0.15387376  0.04971534  0.06391525  0.03185690 -0.02648418 -0.02549678 -0.03718839 -0.07284570 -0.07746888 -0.08746950 -0.10103364
 [56] -0.06566443 -0.07045744 -0.09410074 -0.21918277 -0.21925861 -0.21463217 -0.20175092 -0.24803880 -0.24470177 -0.39440933 -0.17922355
 [67] -0.19665453 -0.28359079 -0.41163641 -0.30560654 -0.60642189 -0.38557830 -0.41624334 -0.60138208 -0.49063715 -0.46777320 -0.77674699
 [78] -0.52893013 -0.62323397 -0.94517291 -0.59306419 -0.86204398 -0.62476873 -0.71131492 -0.88800460 -0.89506322 -0.29657465 -0.64909631
 [89] -0.66163689 -0.35463172 -0.98558891 -0.73017246 -0.99104047 -0.84878296 -1.06225979 -0.53397018 -1.24924064 -1.00359178 -2.65064740
[100] -3.31675935

However, this is what anndata returns. Dimensionality and the shape of the list/array look different.

>>> import anndata
>>> anndata.__version__
'0.11.4'
>>> adata.uns["rank_genes_groups"]["logfoldchanges"]
array([( 1.2091126 ,  1.3671894 ,  1.03600216e+00,  1.25505447e+00,  1.21497381e+00,  0.8379856 ),
       ( 1.0016283 ,  1.520212  ,  8.02816689e-01,  8.07398200e-01,  1.11363661e+00,  1.2618493 ),
       ( 1.2513808 ,  0.9569721 ,  1.06660414e+00,  7.27090359e-01,  7.70881355e-01,  0.6068869 ),
       ( 0.8667367 ,  0.8978347 ,  7.36395359e-01,  6.78247988e-01,  7.51340449e-01,  1.3400772 ),
       ( 0.8735813 ,  0.76945794,  9.31681454e-01,  7.42217124e-01,  1.22592008e+00,  0.8190094 ),
       ( 0.8352304 ,  0.57019335,  6.17068052e-01,  8.51450741e-01,  1.17840135e+00,  0.58131063),
...