add _to_string method in expr

h2o-py/h2o/assembly.py

@@ -70,7 +70,7 @@ def fit(self, fr, **fit_params):
     for step in self.steps:
       res.append(step[1].to_rest(step[0]))
     res = "[" + ",".join([_quoted(r.replace('"',"'")) for r in res]) + "]"
-    j = H2OConnection.post_json(url_suffix="Assembly", steps=res, frame=fr._id, _rest_version=99)
+    j = H2OConnection.post_json(url_suffix="Assembly", steps=res, frame=fr.frame_id, _rest_version=99)
     self.id = j["assembly"]["name"]
     return get_frame(j["result"]["name"])
 

h2o-py/h2o/expr.py

@@ -18,7 +18,7 @@ def __init__(self,op,*ast):
     #       - tuple of Exprs; further: ID is one of
     #       -  - None: Expr is lazy, never evaluated
     #       -  - ""; Expr has been executed once, but no temp ID was made
-    #       -  - String: this Expr is mid-execution, with the given temp ID.  
+    #       -  - String: this Expr is mid-execution, with the given temp ID.
     #            Once execution has completed the ast field will be set to TRUE
     self._ast       = tuple(child._ex if isinstance(child,frame.H2OFrame) else child for child in ast)
     self._id        = None  # "id"  - See above; sometimes None, "", or a "py_tmp" string, or a user string
@@ -117,7 +117,7 @@ def _eval_driver(self,top):
     # temps... to help debug GC issues.
     gc.collect()
     return self
-    
+
   # Magical count-of-5:   (get 2 more when looking at it in debug mode)
   #  2 for _do_it frame, 2 for _do_it local dictionary list, 1 for parent
   MAGIC_REF_COUNT = 5 if sys.gettrace() is None else 7  # M = debug ? 7 : 5
@@ -130,8 +130,8 @@ def _eval_driver(self,top):
   # is eager, so the time parse as to occur in the correct order relative to
   # the timezone change, so cannot be lazy.
   def _do_it(self,top):
-    if self._data is not None:    # Data already computed and cached; could a "false-like" cached value 
-      return self._id if isinstance(self._data,dict) else str(self._data) 
+    if self._data is not None:    # Data already computed and cached; could a "false-like" cached value
+      return self._id if isinstance(self._data,dict) else str(self._data)
     if self._id: return self._id  # Data already computed under ID, but not cached
     # Here self._id is either None or ""
     # Build the eval expression
@@ -159,7 +159,7 @@ def _arg_to_expr(arg):
   def _clear_impl(self):
     if not isinstance(self._ast,tuple): return
     for ast in self._ast:
-      if isinstance(ast,ExprNode): 
+      if isinstance(ast,ExprNode):
         ast._clear_impl()
     if self._id: self._ast = True  # Local pytmp
 
@@ -171,7 +171,7 @@ def __del__(self):
   def _tabulate(self,tablefmt,rollups):
     """
     Pretty tabulated string of all the cached data, and column names
-    """    
+    """
     if not isinstance(self._fetch_data(10),dict):  return str(self._data) # Scalars print normally
     # Pretty print cached data
     d = collections.OrderedDict()
@@ -333,3 +333,6 @@ def _setitem(self, b, c):
     src = c if isinstance(c,ExprNode) else (float("nan") if c is None else c)
     return (ExprNode(":="    ,self,src,col_expr,row_expr) if colname is None
        else ExprNode("append",self,src,colname))
+
+  def _to_string(self):
+    return ' '.join(["("+self._op] + [ExprNode._arg_to_expr(a) for a in self._ast] + [")"])

h2o-py/h2o/frame.py

@@ -1004,7 +1004,7 @@ def hist(self, breaks="Sturges", plot=True, **kwargs):
     total = frame["counts"].sum(True)
     densities = [(frame[i,"counts"]/total)*(1/(frame[i,"breaks"]-frame[i-1,"breaks"])) for i in range(1,frame["counts"].nrow)]
     densities.insert(0,0)
-    densities_frame = H2OFrame.fromPython([[d] for d in densities])
+    densities_frame = H2OFrame.fromPython(densities)
     densities_frame.set_names(["density"])
     frame = frame.cbind(densities_frame)
 

h2o-py/h2o/transforms/preprocessing.py

@@ -91,7 +91,7 @@ def transform(self,X,y=None,**params):
     return X[self.cols]
 
   def to_rest(self, step_name):
-    ast = self._dummy_frame()[self.cols]._ast._debug_print(pprint=False)
+    ast = self._dummy_frame()[self.cols]._ex._to_string()
     return super(H2OColSelect, self).to_rest([step_name,"H2OColSelect",ast,False,"|"])
 
 
@@ -129,7 +129,7 @@ def _transform_helper(self,X,**params):
     return res
 
   def to_rest(self, step_name):
-    ast = self._transform_helper(self._dummy_frame())._ast._debug_print(pprint=False)
+    ast = self._transform_helper(self._dummy_frame())._ex._to_string()
     new_col_names = self.new_col_name
     if new_col_names is None: new_col_names=["|"]
     elif not isinstance(new_col_names, (list,tuple)): new_col_names = [new_col_names]

h2o-py/h2o/transforms/transform_base.py

@@ -1,5 +1,6 @@
 from ..frame import H2OFrame
 import urllib
+from h2o import expr
 
 class TransformAttributeError(AttributeError):
   def __init__(self,obj,method):
@@ -47,9 +48,7 @@ def set_params(self, **params):
 
   @staticmethod
   def _dummy_frame():
-    dummy = H2OFrame()
-    dummy._id = "py_dummy"
-    return dummy
+    return  H2OFrame(expr.ExprNode('dummy'))
 
   def to_rest(self, args):
     return urllib.quote("{}__{}__{}__{}__{}".format(*args))