Nick Pentreath

See https://github.com/tensorflow/models/commit/cce6c09b1c11350f135f07322855f5b9ab7f5b03

Improve API input format

2

Currently the model can only take in a numpy array of weather data, in `txt` format, with a specific hard-coded shape (specifically the number of data rows in the input...

Complete implementation of remaining Spark models

2

## Classifiers - [ ] OneVsRest ## Regressors - [ ] AFTSurvivalRegression - [ ] IsotonicRegression ## Clustering - [ ] BisectingKMeans - [ ] GaussianMixture ## Other - [...

Complete implementation of remaining Spark transformers

3

- [ ] OneHotEncoder - [ ] RFormula - [ ] PolynomialExpansion - [ ] Interaction - [ ] Imputer - [ ] VectorIndexer - [ ] Word2Vec These require...

This throws an exception with the following warning: `org.codehaus.janino.JaninoRuntimeException: Incompatible return types` This can be reproduced with the following code ``` val action = core.eq(inputExpr, "cat") val pfaDoc = new...

[Special forms](http://dmg.org/pfa/docs/special_forms/) include control flow and variable management (cells and pools).

Provide wrappers for `SparkSupport` functionality to provide export of individual transformers/models (as well as pipelines when supported on the Scala side). Use Spark's built-in Py4J support. - [ ] Add...

Currently we assume all columns are set for the various classifiers (`predictionCol`, `rawPredictionCol` and `probabilityCol`). While they are by default and that is definitely the common case, it's not required,...

There are some edge cases in the `project` function for GLM that need to be handled and tested as follow ups.