Fixed dfply documentation on verbs, created longs of those verbs, and improved dflpy import.

grama/__init__.py

@@ -1,81 +1,9 @@
 from .tools import *
 from .core import *
 
-# Integrate dfply tools
+# Import dfply
 # --------------------------------------------------
-from .dfply import Intention, dfdelegate, make_symbolic, convert_type
-from .dfply import var_in, is_nan, not_nan
-from .dfply import starts_with, ends_with, contains, matches, everything
-from .dfply import num_range, one_of, columns_between, columns_from, columns_to
-
-# group.py
-from .dfply import group_by as tf_group_by
-from .dfply import ungroup as tf_ungroup
-
-# join.py
-from .dfply import inner_join as tf_inner_join
-from .dfply import full_join as tf_full_join
-from .dfply import outer_join as tf_outer_join
-from .dfply import left_join as tf_left_join
-from .dfply import right_join as tf_right_join
-from .dfply import semi_join as tf_semi_join
-from .dfply import anti_join as tf_anti_join
-from .dfply import bind_rows as tf_bind_rows
-from .dfply import bind_cols as tf_bind_cols
-
-# reshape.py
-from .dfply import arrange as tf_arrange
-from .dfply import rename as tf_rename
-from .dfply import separate as tf_separate
-from .dfply import unite as tf_unite
-from .dfply import gather as tf_gather
-from .dfply import spread as tf_spread
-from .dfply import explode as tf_explode
-
-# select.py
-from .dfply import select as tf_select
-from .dfply import select_if as tf_select_if
-from .dfply import drop as tf_drop
-from .dfply import drop_if as tf_drop_if
-
-# set_ops.py
-from .dfply import union as tf_union
-from .dfply import intersect as tf_intersect
-from .dfply import set_diff as tf_set_diff
-
-# subset.py
-from .dfply import head as tf_head
-from .dfply import tail as tf_tail
-from .dfply import sample as tf_sample
-from .dfply import distinct as tf_distinct
-from .dfply import row_slice as tf_row_slice
-from .dfply import mask as tf_filter
-from .dfply import top_n as tf_top_n
-from .dfply import pull as tf_pull
-from .dfply import dropna as tf_dropna
-
-# summarize.py
-from .dfply import summarize as tf_summarize
-from .dfply import summarize_each as tf_summarize_each
-
-# summary_functions.py
-from .dfply import mean, first, last, nth, n, n_distinct, IQR, quant
-from .dfply import colmin, colmax, colsum, median, var, sd, binomial_ci
-from .dfply import mse, rmse, ndme, rsq
-from .dfply import corr
-
-# transform.py
-from .dfply import mutate as tf_mutate
-from .dfply import mutate_if as tf_mutate_if
-from .dfply import transmute as tf_transmute
-
-# vector.py
-from .dfply import order_series_by, desc, coalesce, case_when, if_else, na_if
-
-# window_functions.py
-from .dfply import lead, lag, between, dense_rank, min_rank
-from .dfply import cumsum, cummean, cummax, cummin, cumprod, cumany, cumall
-from .dfply import percent_rank, row_number
+from .dfply import *
 
 # Add functionality to dfply
 from .string_helpers import *

grama/comp_building.py

@@ -17,11 +17,12 @@
     "cp_marginals",
 ]
 
-from collections import ChainMap
 import grama as gr
+from collections import ChainMap
 from grama import add_pipe, pipe
-from toolz import curry
 from pandas import concat, DataFrame
+from toolz import curry
+
 
 ## Model Building Interface (MBI) tools
 ##################################################

grama/comp_metamodels.py

@@ -5,6 +5,7 @@
 from grama import add_pipe, pipe
 from toolz import curry
 
+
 ## Fit a metamodel
 # --------------------------------------------------
 @curry

grama/core.py

@@ -16,39 +16,26 @@
     "NaN",
 ]
 
-from abc import ABC, abstractmethod
 import copy
-
-from numpy import (
-    ones,
-    zeros,
-    triu_indices,
-    eye,
-    array,
-    Inf,
-    NaN,
-    sqrt,
-    dot,
-    diag,
-    isfinite,
-)
+import networkx as nx
+import warnings
+import grama as gr
+from grama import pipe, valid_dist, param_dist
+from abc import ABC, abstractmethod
+from itertools import chain
+from numpy import ones, zeros, triu_indices, eye, array, Inf, NaN, sqrt, \
+    dot, diag, isfinite
 from numpy import min as npmin
 from numpy import max as npmax
+from numpy.linalg import cholesky
 from numpy.random import random, multivariate_normal
 from numpy.random import seed as set_seed
+from pandas import DataFrame, concat
 from scipy.linalg import det, LinAlgError, solve
 from scipy.optimize import root_scalar
 from scipy.stats import norm, gaussian_kde
-from pandas import DataFrame, concat
-
-import grama as gr
-from grama import pipe, valid_dist, param_dist
-
-from itertools import chain
-from numpy.linalg import cholesky
 from toolz import curry
-import warnings
-import networkx as nx
+
 
 ## Package settings
 RUNTIME_LOWER = 1  # Cutoff threshold for runtime messages

grama/data/datasets.py

@@ -10,6 +10,7 @@
 from pandas import read_csv
 from pathlib import Path
 
+
 path_this = Path(__file__)
 path_grama = path_this.parents[1]
 

grama/dfply/__init__.py

@@ -4,14 +4,14 @@
 from .mask_helpers import *
 from .reshape import *
 from .select import *
-from .select import *
 from .set_ops import *
 from .subset import *
 from .summarize import *
 from .transform import *
 from .summary_functions import *
-from .window_functions import *
 from .vector import *
+from .window_functions import *
+
 
 for verb in dir():
     if "ize" in verb:

grama/dfply/base.py

@@ -1,9 +1,17 @@
-import pandas as pd
-import numpy as np
+__all__ = [
+    "Intention",
+    "dfdelegate",
+    "make_symbolic",
+    "symbolic_evaluation",
+    "group_delegation",
+    "flatten"
+]
+
 import warnings
+from .. import pipe, add_pipe  # Use grama pipe to preserve metadata
 from functools import partial, wraps
-
-from .. import pipe  # Use grama pipe to preserve metadata
+from numpy import zeros, array
+from pandas import Series, Index, DataFrame
 
 
 def _recursive_apply(f, l):
@@ -12,8 +20,7 @@ def _recursive_apply(f, l):
         if isinstance(l, tuple):
             out = tuple(out)
         return out
-    else:
-        return f(l)
+    return f(l)
 
 
 def contextualize(arg, context):
@@ -60,8 +67,7 @@ def wrapper(*args, **kwargs):
         if delay:
             delayed = _delayed_function(f, args, kwargs)
             return Intention(delayed)
-        else:
-            return f(*args, **kwargs)
+        return f(*args, **kwargs)
 
     ## Preserve documentation
     wrapper.__doc__ = f.__doc__
@@ -227,9 +233,9 @@ def _evaluate_label(self, df, arg):
         arg = self._evaluate(df, arg)
 
         cols = list(df.columns)
-        if isinstance(arg, pd.Series):
+        if isinstance(arg, Series):
             arg = arg.name
-        if isinstance(arg, pd.Index):
+        if isinstance(arg, Index):
             arg = list(arg)
         if isinstance(arg, int):
             arg = cols[arg]
@@ -242,11 +248,11 @@ def _evaluate_selector(self, df, arg):
             arg = arg.evaluate(df)
 
         cols = list(df.columns)
-        if isinstance(arg, pd.Series):
+        if isinstance(arg, Series):
             arg = [cols.index(arg.name)]
-        if isinstance(arg, pd.Index):
+        if isinstance(arg, Index):
             arg = [cols.index(i) for i in list(arg)]
-        if isinstance(arg, pd.DataFrame):
+        if isinstance(arg, DataFrame):
             arg = [cols.index(i) for i in arg.columns]
         if isinstance(arg, int):
             arg = [arg]
@@ -255,8 +261,8 @@ def _evaluate_selector(self, df, arg):
         if isinstance(arg, (list, tuple)):
             arg = [cols.index(i) if isinstance(i, str) else i for i in arg]
 
-        selection_vector = np.zeros(df.shape[1])
-        col_idx = np.array(arg)
+        selection_vector = zeros(df.shape[1])
+        col_idx = array(arg)
 
         if negate and len(col_idx) > 0:
             selection_vector[col_idx] = -1
@@ -267,8 +273,7 @@ def _evaluate_selector(self, df, arg):
     def _evaluator_loop(self, df, arg, eval_func):
         if isinstance(arg, (list, tuple)):
             return [self._evaluator_loop(df, a_, eval_func) for a_ in arg]
-        else:
-            return eval_func(df, arg)
+        return eval_func(df, arg)
 
     def _symbolic_eval(self, df, arg):
         return self._evaluator_loop(df, arg, self._evaluate)
@@ -314,16 +319,16 @@ def _recursive_kwarg_eval(self, df, kwargs):
         }
 
     def _find_eval_args(self, request, args):
-        if (request == True) or ("*" in request):
+        if (request is True) or ("*" in request):
             return [i for i in range(len(args))]
-        elif request in [None, False]:
+        if request in [None, False]:
             return []
         return request
 
     def _find_eval_kwargs(self, request, kwargs):
-        if (request == True) or ("**" in request):
+        if (request is True) or ("**" in request):
             return [k for k in kwargs.keys()]
-        elif request in [None, False]:
+        if request in [None, False]:
             return []
         return request
 
@@ -341,21 +346,21 @@ def symbolic_evaluation(
 ):
     if function:
         return IntentionEvaluator(function)
-    else:
-
-        @wraps(function)
-        def wrapper(function):
-            return IntentionEvaluator(
-                function,
-                eval_symbols=eval_symbols,
-                eval_as_label=eval_as_label,
-                eval_as_selector=eval_as_selector,
-            )
 
-        return wrapper
+    @wraps(function)
+    def wrapper(function):
+        return IntentionEvaluator(
+            function,
+            eval_symbols=eval_symbols,
+            eval_as_label=eval_as_label,
+            eval_as_selector=eval_as_selector,
+        )
+
+    return wrapper
 
 
 class group_delegation(object):
+
     __name__ = "group_delegation"
 
     def __init__(self, function):
@@ -386,19 +391,21 @@ def __call__(self, *args, **kwargs):
         grouped_by = getattr(args[0], "_grouped_by", None)
         if (grouped_by is None) or not all([g in args[0].columns for g in grouped_by]):
             return self.function(*args, **kwargs)
-        else:
-            applied = self._apply(args[0], *args[1:], **kwargs)
 
-            with warnings.catch_warnings():
-                warnings.simplefilter("ignore")
-                applied._grouped_by = grouped_by
+        applied = self._apply(args[0], *args[1:], **kwargs)
+
+        with warnings.catch_warnings():
+            warnings.simplefilter("ignore")
+            applied._grouped_by = grouped_by
 
-            return applied
+        return applied
 
 
 def dfpipe(f):
-    return pipe(group_delegation(symbolic_evaluation(f)))
+    return add_pipe(group_delegation(symbolic_evaluation(f)))
 
 
 def dfdelegate(f):
-    return group_delegation(symbolic_evaluation(f))
+    class addName(group_delegation):
+        __name__ = f.__name__
+    return addName(group_delegation(symbolic_evaluation(f)))

grama/dfply/group.py

@@ -1,14 +1,24 @@
-from .base import *
+__all__ = [
+    "tran_group_by",
+    "tf_group_by",
+    "tran_ungroup",
+    "tf_ungroup",
+]
+
+from .base import symbolic_evaluation
+from .. import add_pipe
 
 
-@pipe
 @symbolic_evaluation(eval_as_label=True)
-def group_by(df, *args):
+def tran_group_by(df, *args):
     df._grouped_by = list(args)
     return df
 
+tf_group_by = add_pipe(tran_group_by)
+
 
-@pipe
-def ungroup(df):
+def tran_ungroup(df):
     df._grouped_by = None
     return df
+
+tf_ungroup = add_pipe(tran_ungroup)