rdkit build fail

[mark-str]

Trying to build RDKit and have got down to one failed test from six (improvement from adding pandas<0.2 to meta.yaml and using a Python 5.3 environment), but still have one failed test #111 'pythonTestDirML'.

Can anyone tell me how to solve it?

Currently using: Python 3.5.4 :: Anaconda custom (64-bit) conda-build 3.2.2 Ubuntu 16.04.3 LTS

Test output (all tests prior, pass):

### ```
108/116 Test #111: pythonTestDirML ..................***Failed   12.64 sec
.......
----------------------------------------------------------------------
Ran 7 tests in 0.681s
OK
............
----------------------------------------------------------------------
Ran 12 tests in 0.308s
OK
F
======================================================================
FAIL: test1_Issue163 (__main__.TestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "UnitTestAnalyzeComposite.py", line 59, in test1_Issue163
    self.assertTrue(res[0][0] == 'BALABANJ')
AssertionError: False is not true
----------------------------------------------------------------------
Ran 1 test in 0.043s
FAILED (failures=1)
...
----------------------------------------------------------------------
Ran 3 tests in 0.001s
OK
.......
----------------------------------------------------------------------
Ran 7 tests in 0.001s
OK
...
----------------------------------------------------------------------
Ran 3 tests in 0.353s
OK
....
----------------------------------------------------------------------
Ran 4 tests in 0.003s
OK
/home/marks/projects/MDC/anaconda3/conda-bld/rdkit_1516631519704/work/rdkit/ML/Data/DataUtils.py:199: FutureWarning: split() requires a non-empty pattern match.
  resArr = expr2.split(inLine)[:-1]
.../home/marks/projects/MDC/anaconda3/conda-bld/rdkit_1516631519704/work/rdkit/ML/Data/DataUtils.py:163: FutureWarning: split() requires a non-empty pattern match.
  resArr = expr2.split(inLine)
...
----------------------------------------------------------------------
Ran 6 tests in 0.004s
OK
............
----------------------------------------------------------------------
Ran 12 tests in 0.002s
OK
.....
----------------------------------------------------------------------
Ran 5 tests in 0.034s
OK
........
----------------------------------------------------------------------
Ran 8 tests in 0.001s
OK
testing AddChild: 
testing GetName: 
testing GetChildren: 
testing PruneChild: 
testing tree equals: 
testing pickled tree equals: 
testing deepcopy on trees: 
testing list membership: ...
----------------------------------------------------------------------
Ran 3 tests in 0.103s
OK
use selection with replacement: 
test the results of CrossValidation: 
test that the CrossValidationDriver runs: ......
----------------------------------------------------------------------
Ran 6 tests in 0.019s
OK
testing basic tree growth: 
testing basic tree classification: 
testing multivalued tree growth: 
testing basic tree growth (python entropy code): 
testing tree classification (python entropy code): 
testing multivalued tree growth (python entropy code): .
----------------------------------------------------------------------
Ran 1 test in 0.007s
OK
.........
----------------------------------------------------------------------
Ran 9 tests in 0.014s
OK
testing tree comparisons: 
testing tree1: 
testing tree2: 
testing classification: 
testing unused variables: 
a more extensive test of the cmp stuff using hand-built trees: 
a more extensive test of the cmp stuff using pickled trees: 
try recycling descriptors: 
try random forests descriptors: ....
----------------------------------------------------------------------
Ran 4 tests in 0.107s
OK
..
----------------------------------------------------------------------
Ran 2 tests in 0.002s
OK
testing calculation of a single descriptor: 
testing calculation of multiple descriptors: ...
----------------------------------------------------------------------
Ran 3 tests in 0.009s
OK
..
----------------------------------------------------------------------
Ran 2 tests in 0.006s
OK
..
----------------------------------------------------------------------
Ran 2 tests in 0.299s
OK
....
----------------------------------------------------------------------
Ran 4 tests in 0.580s
OK
.....
----------------------------------------------------------------------
Ran 5 tests in 0.256s
OK
....
----------------------------------------------------------------------
Ran 4 tests in 0.010s
OK
...
----------------------------------------------------------------------
Ran 3 tests in 0.552s
OK
testing backprop training on and: 
testing backprop training on a linear function: 
testing backprop training on or: ..
----------------------------------------------------------------------
Ran 2 tests in 0.000s
OK

testing Burges empirical risk bound: 
testing Cherkassky empirical risk bound: !!! TEST FAILURE:  python UnitTestAnalyzeComposite.py {}

        Start 112: pythonTestDirDataStructs
109/116 Test  #83: moldraw2DTest1 ...................   Passed   41.63 sec
        Start 113: pythonTestDirDbase
110/116 Test #113: pythonTestDirDbase ...............   Passed    1.21 sec
        Start 114: pythonTestDirSimDivFilters
111/116 Test #112: pythonTestDirDataStructs .........   Passed    2.11 sec
        Start 115: pythonTestDirVLib
112/116 Test #114: pythonTestDirSimDivFilters .......   Passed    0.70 sec
        Start 116: pythonTestDirChem
113/116 Test #115: pythonTestDirVLib ................   Passed    3.21 sec
114/116 Test #110: pythonTestDbCLI ..................   Passed   23.57 sec
115/116 Test  #75: testMolAlign .....................   Passed  116.37 sec
116/116 Test #116: pythonTestDirChem ................   Passed   62.71 sec

99% tests passed, 1 tests failed out of 116

Total Test time (real) = 202.15 sec

The following tests FAILED:
        111 - pythonTestDirML (Failed)
Errors while running CTest