[SW-2567] Fix CoxPH docs example for Scala and Python

[neema-m]

Fixes JIRA: https://h2oai.atlassian.net/browse/SW-2567

Current example fails for

org.apache.spark.sql.AnalysisException: cannot resolve '`label`' given input columns: [start, transplant, stop, age, event, year, id, surgery];;
'Project [start#2305, stop#2306, event#2307, age#2308, year#2309, surgery#2310, transplant#2311, id#2312, 'label]
+- Relation[start#2305,stop#2306,event#2307,age#2308,year#2309,surgery#2310,transplant#2311,id#2312] csv

need to add parameters that are needed for CoxPH

[h2o-ops]

Hi @neemah2o, your PR title "Create sw_coxph.rst" is missing JIRA issue number! Please specify it in the following form [SW-]

[mn-mikke]

It seems we need to get a different dataset with test data, all predictions are NaN:

+------------------+-----------------+-------+----------+-------------------+----------+
|age               |year             |surgery|transplant|detailed_prediction|prediction|
+------------------+-----------------+-------+----------+-------------------+----------+
|-12.9390828199863 |6.39561943874059 |1      |0         |{NaN}              |NaN       |
|-0.249144421629019|5.77686516084873 |0      |1         |{NaN}              |NaN       |
|3.04722792607803  |5.51129363449692 |0      |1         |{NaN}              |NaN       |
|-15.3401779603012 |4.19712525667351 |0      |1         |{NaN}              |NaN       |
|0.0164271047227942|3.97809719370294 |1      |0         |{NaN}              |NaN       |
|-0.208076659822041|3.75085557837098 |0      |0         |{NaN}              |NaN       |
|-4.53388090349076 |2.30800821355236 |0      |0         |{NaN}              |NaN       |
|0.610540725530456 |3.27720739219713 |1      |1         |{NaN}              |NaN       |
|-1.86817          |4.966785763      |1      |1         |{NaN}              |NaN       |
|13.5003422313484  |2.56536618754278 |0      |0         |{NaN}              |NaN       |
|-17.4647501711157 |1.58247775496235 |0      |0         |{NaN}              |NaN       |
|3.80013689253936  |1.56605065023956 |0      |0         |{NaN}              |NaN       |
|-11.8165639972622 |3.26351813826146 |0      |1         |{NaN}              |NaN       |
|1.44832306639288  |1.07049965776865 |0      |0         |{NaN}              |NaN       |
|6.59548254620123  |0.700889801505818|0      |0         |{NaN}              |NaN       |
|2.51882272416153  |2.63381245722108 |0      |1         |{NaN}              |NaN       |
|4.03285420944559  |5.51403148528405 |1      |1         |{NaN}              |NaN       |
|0.703627652292951 |5.09240246406571 |0      |1         |{NaN}              |NaN       |
|-4.15879534565367 |5.95482546201232 |1      |0         |{NaN}              |NaN       |
|0.03490759753595  |2.157401         |0      |0         |{NaN}              |NaN       |
+------------------+-----------------+-------+----------+-------------------+----------+

Testing dataset doesn't have start and stop column. https://raw.githubusercontent.com/h2oai/sparkling-water/master/examples/smalldata/coxph_test/heart_test.csv Is it expected?

[neema-m]

Good point @mn-mikke , the H2O 3 version uses heart.csv and splits it

heart = h2o.import_file("http://s3.amazonaws.com/h2o-public-test-data/smalldata/coxph_test/heart.csv")

# Split the dataset into a train and test set:
train, test = heart.split_frame(ratios = [.8], seed = 1234)

I will modify the PR to match it

[neema-m]

@mn-mikke , I made the changes to give a proper test dataset:

scala> model.transform(testingDF).show(false)
+-----+----+-----+------------------+-----------------+-------+----------+---+----------------------+--------------------+
|start|stop|event|age               |year             |surgery|transplant|id |detailed_prediction   |prediction          |
+-----+----+-----+------------------+-----------------+-------+----------+---+----------------------+--------------------+
|0    |2   |0    |0.610540725530456 |3.27720739219713 |1      |0         |46 |[-0.45681461656609834]|-0.45681461656609834|
|0    |2   |1    |4.18069815195072  |4.99657768651609 |0      |0         |75 |[0.03651537014229078] |0.03651537014229078 |
|0    |2   |1    |4.56399726214921  |4.17522245037645 |0      |0         |61 |[0.16226172474195466] |0.16226172474195466 |
|0    |4   |0    |-21.2731006160164 |4.87063655030801 |0      |0         |72 |[-0.5015662484269671] |-0.5015662484269671 |
|0    |4   |0    |-6.52977412731006 |2.5927446954141  |0      |0         |38 |[0.20394003741166356] |0.20394003741166356 |
|0    |4   |0    |-4.15879534565367 |5.95482546201232 |1      |0         |94 |[-0.9108067656338691] |-0.9108067656338691 |
|0    |13  |0    |-21.3497604380561 |6.00958247775496 |0      |0         |96 |[-0.6054142887730242] |-0.6054142887730242 |
|0    |18  |0    |7.27994524298425  |1.3305954825462  |0      |0         |20 |[0.7266981608568234]  |0.7266981608568234  |
|0    |19  |0    |13.5003422313484  |2.56536618754278 |0      |0         |37 |[0.6295012210602893]  |0.6295012210602893  |
|0    |26  |0    |-1.55509924709104 |5.18275154004107 |1      |0         |80 |[-0.7557212094322586] |-0.7557212094322586 |
|0    |28  |0    |1.44832306639288  |1.07049965776865 |0      |0         |16 |[0.6448465982857511]  |0.6448465982857511  |
|0    |36  |0    |-7.73716632443532 |0.490075290896646|0      |0         |4  |[0.5033912420589028]  |0.5033912420589028  |
|0    |40  |1    |-5.42094455852156 |3.16495550992471 |1      |0         |44 |[-0.570992880113988]  |-0.570992880113988  |
|0    |41  |0    |-0.657084188911703|3.75085557837098 |0      |0         |53 |[0.10108321544611465] |0.10108321544611465 |
|0    |46  |0    |0.925393566050651 |2.507871321013   |0      |0         |36 |[0.36831841133591947] |0.36831841133591947 |
|0    |51  |0    |0.903490759753595 |2.1574264202601  |0      |0         |33 |[0.45740528733585895] |0.45740528733585895 |
|0    |51  |0    |2.86926762491444  |0.780287474332649|0      |0         |7  |[0.663074262583844]   |0.663074262583844   |
|0    |78  |0    |-6.61738535249829 |3.99452429842574 |1      |0         |59 |[-0.703921414261432]  |-0.703921414261432  |
|0    |102 |1    |-6.75154004106776 |3.564681724846   |0      |0         |52 |[0.0260867300031159]  |0.0260867300031159  |
|0    |428 |0    |-18.798083504449  |4.08487337440109 |0      |0         |82 |[-0.02612265146660009]|-0.02612265146660009|
+-----+----+-----+------------------+-----------------+-------+----------+---+----------------------+--------------------+
only showing top 20 rows

[mn-mikke]

LGTM. Thank you @neemah2o!