Renew feature (#3416)

* to fix issue 3412#

* to fix issue 3412#

* to fix the issue 3412#

* to fix issue 3412

* fix the issue 3412

* to fix issue 3412

* to fix the issue 3412#

* 3412#

* to fix issue 3412#

* changes suggested by junpenglao

* minor

Author: chang111

pymc3/diagnostics.py

@@ -4,6 +4,7 @@
 from .stats import statfunc, autocov
 from .util import get_default_varnames
 from .backends.base import MultiTrace
+import warnings
 
 __all__ = ['geweke', 'gelman_rubin', 'effective_n']
 
@@ -97,7 +98,8 @@ def geweke(x, first=.1, last=.5, intervals=20):
         return np.array(zscores)
 
 
-def gelman_rubin(mtrace, varnames=None, include_transformed=False):
+
+def gelman_rubin(mtrace, var_names=None, include_transformed=False, **kwargs):
     R"""Returns estimate of R for a set of traces.
 
     The Gelman-Rubin diagnostic tests for lack of convergence by comparing
@@ -141,7 +143,12 @@ def gelman_rubin(mtrace, varnames=None, include_transformed=False):
     ----------
     Brooks and Gelman (1998)
     Gelman and Rubin (1992)"""
-
+    if 'varnames' in kwargs:
+        var_names = kwargs['varnames']
+        warnings.warn(
+            'Keyword argument varnames renamed to var_names, and will be removed in pymc3 3.8',
+            DeprecationWarning
+            )
     def rscore(x, num_samples):
         # Calculate between-chain variance
         B = num_samples * np.var(np.mean(x, axis=1), axis=0, ddof=1)
@@ -163,20 +170,20 @@ def rscore(x, num_samples):
             'Gelman-Rubin diagnostic requires multiple chains '
             'of the same length.')
 
-    if varnames is None:
-        varnames = get_default_varnames(mtrace.varnames, include_transformed=include_transformed)
+    if var_names is None:
+        var_names = get_default_varnames(mtrace.varnames, include_transformed=include_transformed)
 
     Rhat = {}
 
-    for var in varnames:
+    for var in var_names:
         x = np.array(mtrace.get_values(var, combine=False))
         num_samples = x.shape[1]
         Rhat[var] = rscore(x, num_samples)
 
     return Rhat
 
 
-def effective_n(mtrace, varnames=None, include_transformed=False):
+def effective_n(mtrace, var_names=None, include_transformed=False, **kwargs):
     R"""Returns estimate of the effective sample size of a set of traces.
 
     Parameters
@@ -211,7 +218,11 @@ def effective_n(mtrace, varnames=None, include_transformed=False):
     References
     ----------
     Gelman et al. BDA (2014)"""
-
+    if 'varnames' in kwargs:
+        var_names = kwargs['varnames']
+        warnings.warn(
+            'Keyword argument varnames renamed to var_names, and will be removed in pymc3 3.8',
+            DeprecationWarning)
     def get_neff(x):
         """Compute the effective sample size for a 2D array
         """
@@ -291,12 +302,12 @@ def generate_neff(trace_values):
             'Calculation of effective sample size requires multiple chains '
             'of the same length.')
 
-    if varnames is None:
-        varnames = get_default_varnames(mtrace.varnames, include_transformed=include_transformed)
+    if var_names is None:
+        var_names = get_default_varnames(mtrace.varnames, include_transformed=include_transformed)
 
     n_eff = {}
 
-    for var in varnames:
+    for var in var_names:
         n_eff[var] = generate_neff(mtrace.get_values(var, combine=False))
 
     return n_eff

pymc3/stats.py

@@ -16,6 +16,7 @@
 import pymc3 as pm
 from pymc3.theanof import floatX
 
+
 if pkg_resources.get_distribution('scipy').version < '1.0.0':
     from scipy.misc import logsumexp
 else:
@@ -850,9 +851,11 @@ def dict2pd(statdict, labelname):
     statpd = statpd.rename(labelname)
     return statpd
 
-def summary(trace, varnames=None, transform=lambda x: x, stat_funcs=None,
+
+
+def summary(trace, var_names=None, transform=lambda x: x, stat_funcs=None,
                extend=False, include_transformed=False,
-               alpha=0.05, start=0, batches=None):
+               alpha=0.05, start=0, batches=None, **kwargs):
     R"""Create a data frame with summary statistics.
 
     Parameters
@@ -936,10 +939,16 @@ def summary(trace, varnames=None, transform=lambda x: x, stat_funcs=None,
         mu__0  0.066473  0.000312  0.105039  0.214242
         mu__1  0.067513 -0.159097 -0.045637  0.062912
     """
+    if 'varnames' in kwargs:
+        var_names = kwargs['varnames']
+        warnings.warn(
+            'Keyword argument varnames renamed to var_names, and will be removed in pymc3 3.8',
+            DeprecationWarning
+            )
     from .backends import tracetab as ttab
 
-    if varnames is None:
-        varnames = get_default_varnames(trace.varnames,
+    if var_names is None:
+        var_names = get_default_varnames(trace.varnames,
                        include_transformed=include_transformed)
 
     if batches is None:
@@ -957,7 +966,7 @@ def summary(trace, varnames=None, transform=lambda x: x, stat_funcs=None,
             funcs = stat_funcs
 
     var_dfs = []
-    for var in varnames:
+    for var in var_names:
         vals = transform(trace.get_values(var, burn=start, combine=True))
         flat_vals = vals.reshape(vals.shape[0], -1)
         var_df = pd.concat([f(flat_vals) for f in funcs], axis=1)
@@ -971,11 +980,11 @@ def summary(trace, varnames=None, transform=lambda x: x, stat_funcs=None,
         return dforg
     else:
         n_eff = pm.effective_n(trace,
-                               varnames=varnames,
+                               varnames=var_names,
                                include_transformed=include_transformed)
         n_eff_pd = dict2pd(n_eff, 'n_eff')
         rhat = pm.gelman_rubin(trace,
-                               varnames=varnames,
+                               varnames=var_names,
                                include_transformed=include_transformed)
         rhat_pd = dict2pd(rhat, 'Rhat')
         return pd.concat([dforg, n_eff_pd, rhat_pd],