WiP: Fancy print

[dinga92]

Nice printing of dataset objects, so that samples and sa are in rows and features/fa are in columns. Just disclosing it here for a feedback, not intended for merge at the moment. For now it is using external library "tabulate" (pipable, MIT licence).

example:
ds = normal_feature_dataset(nonbogus_features=[0,1])
ds.prt()

------  -------  ----------------  -----  -----  -----  -----
                 nonbogus_targets  L0     L1
chunks  targets
0       L0                         0.130  -0.10  -0.02  0.077
0       L0                         0.318  0.238  -0.14  -0.40
0       L0                         0.286  -0.02  0.101  -0.24
0       L0                         0.757  -0.20  -0.43  -0.21
0       L0                         0.345  0.070  -0.31  0.010
0       L0                         0.613  0.217  0.125  -0.32
0       L0                         0.329  0.129  -0.20  0.024
0       L0                         0.691  -0.06  0.074  0.375
0       L0                         0.745  -0.03  0.173  0.295
0       L0                         0.282  -0.23  0.188  0.281
1       L0                         0.244  -0.26  0.042  0.171
1       L0                         0.485  -0.10  -0.35  0.319
1       L0                         0.715  -0.10  0.103  0.178
1       L0                         0.710  -0.04  0.016  0.337
1       L0                         0.586  0.225  -0.37  0.085
1       L0                         0.633  -0.07  -0.23  0.369
1       L0                         0.054  -0.47  -0.20  0.306
1       L0                         0.655  -0.04  -0.19  0.095
1       L0                         0.648  -0.09  0.261  -0.26
1       L0                         0.659  -0.40  -0.03  0.350
2       L0                         0.152  -0.00  0.133  0.221
2       L0                         0.611  0.121  0.273  -0.35
2       L0                         1.0    0.119  -0.13  -0.21
2       L0                         0.118  -0.30  -0.10  -0.04
2       L0                         0.467  -0.16  0.723  0.048
2       L0                         0.375  -0.16  0.389  -0.19
2       L0                         0.192  0.201  0.129  0.441
2       L0                         0.308  -0.03  0.096  0.161
2       L0                         0.426  0.177  -0.03  -0.42
2       L0                         0.293  -0.59  0.217  0.203
3       L0                         0.526  -0.23  -0.28  0.089
3       L0                         0.778  -0.41  -0.16  -0.11
3       L0                         0.457  -0.06  0.552  -0.29
3       L0                         0.412  -0.06  0.286  -0.11
3       L0                         0.434  0.093  -0.07  -0.00
3       L0                         0.457  0.200  0.004  -0.31
3       L0                         0.146  -0.01  0.119  0.086
3       L0                         0.589  -0.07  -0.17  -0.53
3       L0                         0.741  -0.16  -0.25  -0.13
3       L0                         0.803  -0.01  0.205  -0.00
4       L0                         0.593  0.201  0.227  -0.25
4       L0                         0.378  -0.21  -0.09  0.325
4       L0                         0.473  -0.11  -0.45  -0.01
4       L0                         0.322  0.206  0.119  0.394
4       L0                         0.785  -0.15  0.190  0.139
4       L0                         0.317  0.211  -0.19  0.149
4       L0                         0.406  0.162  0.037  0.259
4       L0                         0.739  0.079  -0.16  -0.32
4       L0                         0.317  0.363  -0.36  -0.07
4       L0                         0.486  0.143  -0.00  0.019
------  -------  ----------------  -----  -----  -----  -----
Ommitted 50 rows and 0 columns

ds.prt(sa='targets', height=10)

-------  ----------------  -----  -----  -----  -----
         nonbogus_targets  L0     L1
targets
L0                         0.531  0.254  0.227  0.131
L0                         0.664  -0.02  0.117  -0.13
L0                         0.501  0.157  0.039  -0.29
L0                         0.250  0.237  0.122  0.034
L0                         0.617  -0.09  -0.13  0.419
L0                         0.105  0.033  -0.04  -0.40
L0                         0.167  0.040  0.136  0.186
L0                         0.578  -0.08  -0.20  0.656
L0                         0.154  0.032  -0.17  0.320
L0                         0.129  -0.03  0.335  0.144
-------  ----------------  -----  -----  -----  -----
Ommitted 90 rows and 0 columns

[yarikoptic]

in general i wholeheartedly support this feature! Could you include example in the pr description of how it would look like?

[dinga92]

of course I immediately found a bug and it doesn't work for datasets with attributesCollection :)

[yarikoptic]

would be nice also to add a smoke test which would sweep through all test datasets and see if code doesn't blow. smth like

@sweepargs(ds=datasets.itervalues())
def test_to_table_test_datasets(ds):
    dstxt = ds.to_table()
    for a in itertools.chain(ds.sa.keys(), ds.fa.keys(), ds.a.keys()):
         assert(a in dstxt)

[yarikoptic]

@dinga92 interested to bring this baby to the finish line? ;)

[mih]

@dinga92 Ping! Let's get this done -- would be most appreciated!