fix: fix load single peak in py3 support

news/fix-single-peak.rst

@@ -0,0 +1,23 @@
+**Added:**
+
+* <news item>
+
+**Changed:**
+
+* <news item>
+
+**Deprecated:**
+
+* <news item>
+
+**Removed:**
+
+* <news item>
+
+**Fixed:**
+
+* Fixed extracting single peak with `py2` legacy cleanup
+
+**Security:**
+
+* <news item>

src/diffpy/srmise/applications/extract.py

@@ -18,6 +18,21 @@
 import numpy as np
 
 
+def _baseline_namespace():
+    """Return the baseline classes supported by the CLI."""
+    from diffpy.srmise.baselines.arbitrary import Arbitrary
+    from diffpy.srmise.baselines.fromsequence import FromSequence
+    from diffpy.srmise.baselines.nanospherical import NanoSpherical
+    from diffpy.srmise.baselines.polynomial import Polynomial
+
+    return {
+        "Arbitrary": Arbitrary,
+        "FromSequence": FromSequence,
+        "NanoSpherical": NanoSpherical,
+        "Polynomial": Polynomial,
+    }
+
+
 def main():
     """Default SrMise entry-point."""
 
@@ -483,10 +498,13 @@ def main():
 
         bl = NanoSpherical()
         options.baseline = parsepars(bl, options.bspherical)
-
+    elif options.baseline is not None:
         try:
-            options.baseline = eval("baselines." + options.baseline)
-
+            options.baseline = eval(
+                options.baseline,
+                {"__builtins__": {}},
+                _baseline_namespace(),
+            )
         except Exception as err:
             print(err)
             print("Could not create baseline '%s'. Exiting." % options.baseline)

src/diffpy/srmise/dataclusters.py

@@ -221,7 +221,7 @@ def __next__(self):
                 self.lastcluster_idx = nearest_cluster[0] + 1
             self.clusters = np.insert(
                 self.clusters,
-                int(self.lastcluster_idx),
+                self.lastcluster_idx,
                 [test_idx, test_idx],
                 0,
             )
@@ -289,7 +289,7 @@ def find_nearest_cluster(self, idx):
             return None
 
         flat_idx = clusters_flat.searchsorted(idx)
-        near_idx = flat_idx / 2
+        near_idx = flat_idx // 2
 
         if flat_idx == len(clusters_flat):
             # test_idx is right of the last cluster
@@ -304,8 +304,8 @@ def find_nearest_cluster(self, idx):
         # Calculate which of the two nearest clusters is closer
         distances = np.array(
             [
-                self.x[idx] - self.x[self.clusters[int(near_idx) - 1, 1]],
-                self.x[idx] - self.x[self.clusters[int(near_idx), 0]],
+                self.x[idx] - self.x[self.clusters[near_idx - 1, 1]],
+                self.x[idx] - self.x[self.clusters[near_idx, 0]],
             ]
         )
         if distances[0] < np.abs(distances[1]):

src/diffpy/srmise/modelcluster.py

@@ -1174,7 +1174,7 @@ def cleanfit(self):
         pos = np.array([p["position"] for p in self.model])
         left_idx = pos.searchsorted(self.r_cluster[0])
         right_idx = pos.searchsorted(self.r_cluster[-1])
-        outside_idx = range(0, left_idx)
+        outside_idx = list(range(0, left_idx))
         outside_idx.extend(range(right_idx, len(self.model)))
         # inside_idx = range(left_idx, right_idx)
 

src/diffpy/srmise/modelevaluators/aic.py

@@ -85,7 +85,7 @@ def evaluate(self, fit, count_fixed=False, kshift=0):
         if self.chisq is None:
             self.chisq = self.chi_squared(fit.value(), fit.y_cluster, fit.error_cluster)
 
-        self.stat = self.chisq + self.parpenalty(k, n)
+        self.stat = self.chisq + self.parpenalty(k)
 
         return self.stat
 
@@ -169,7 +169,7 @@ def growth_justified(self, fit, k_prime):
             logger.warning("AIC.growth_justified(): too few data to evaluate quality reliably.")
             n = self.minpoints(k_actual)
 
-        penalty = self.parpenalty(k_test, n) - self.parpenalty(k_actual, n)
+        penalty = self.parpenalty(k_test) - self.parpenalty(k_actual)
 
         return penalty < self.chisq
 

src/diffpy/srmise/modelevaluators/base.py

@@ -76,7 +76,7 @@ def __lt__(self, other):
         assert self.method == other.method  # Comparison between same types required
         assert self.stat is not None and other.stat is not None  # The statistic must already be calculated
 
-        if self.higher_is_better is not None:
+        if self.higher_is_better:
             return self.stat < other.stat
         else:
             return other.stat < self.stat
@@ -87,7 +87,7 @@ def __le__(self, other):
         assert self.method == other.method  # Comparison between same types required
         assert self.stat is not None and other.stat is not None  # The statistic must already be calculated
 
-        if self.higher_is_better is not None:
+        if self.higher_is_better:
             return self.stat <= other.stat
         else:
             return other.stat <= self.stat
@@ -114,7 +114,7 @@ def __gt__(self, other):
         assert self.method == other.method  # Comparison between same types required
         assert self.stat is not None and other.stat is not None  # The statistic must already be calculated
 
-        if self.higher_is_better is not None:
+        if self.higher_is_better:
             return self.stat > other.stat
         else:
             return other.stat > self.stat
@@ -125,7 +125,7 @@ def __ge__(self, other):
         assert self.method == other.method  # Comparison between same types required
         assert self.stat is not None and other.stat is not None  # The statistic must already be calculated
 
-        if self.higher_is_better is not None:
+        if self.higher_is_better:
             return self.stat >= other.stat
         else:
             return other.stat >= self.stat

src/diffpy/srmise/modelparts.py

@@ -430,8 +430,9 @@ def __getitem__(self, index):
         if isinstance(index, tuple) and len(index) == 2:
             start, end = index
             return self.__class__(super().__getitem__(slice(start, end)))
-        else:
-            return super().__getitem__(index)
+        if isinstance(index, slice):
+            return self.__class__(super().__getitem__(index))
+        return super().__getitem__(index)
 
     def transform(self, in_format="internal", out_format="internal"):
         """Transforms format of parameters in this modelpart.

src/diffpy/srmise/pdfpeakextraction.py

@@ -247,7 +247,7 @@ def defaultvars(self, *args):
 
         # Enable "dg" as alias for "effective_dy"
         if "dg" in args and "effective_dy" not in args:
-            nargs.add("effective_dy")
+            nargs.append("effective_dy")
 
         # Set other defaults
         PeakExtraction.defaultvars(self, *nargs)
@@ -888,9 +888,9 @@ def writepwastr(self, comments):
             # Generate parameter labels from the baseline function's parameterdict
             blf = self.extracted.baseline.owner()
             if blf.npars > 0:
-                parlbl = blf.parameterdict.keys()
-                paridx = np.array(blf.parameterdict.values()).argsort()
-                lines.append("# " + " ".join([str(parlbl[i]) for i in paridx]))
+                parlbl = list(blf.parameterdict.keys())
+                paridx = np.array(list(blf.parameterdict.values())).argsort()
+                lines.append("# " + " ".join(str(parlbl[i]) for i in paridx))
                 blpars = " ".join([str(p) for p in self.extracted.baseline.pars])
             else:
                 blpars = "(no parameters)"
@@ -1000,7 +1000,7 @@ def find_qmax(r, y, showgraphs=False):
     new_y = resample(r, y, new_r)
     new_dr = (new_r[-1] - r[0]) / (len(new_r) - 1)
 
-    yfft = np.imag(np.fft.fft(new_y))[: len(new_y) / 2]
+    yfft = np.imag(np.fft.fft(new_y))[: len(new_y) // 2]
 
     d_ratio = stdratio(yfft)
 

src/diffpy/srmise/peakextraction.py

@@ -969,7 +969,7 @@ def extract_single(self, recursion_depth=1):
                 # near_peaks: array containing the indices of two nearest peaks on either side of border_x
                 # other_peaks: all the other peaks in full_cluster
                 # left_data, right_data: indices defining the extent of the "interpeak range" for x, etc.
-                near_peaks = np.array([], dtype=np.int)
+                near_peaks = np.array([], dtype=np.int64)
 
                 # interpeak range goes from peak to peak of next nearest peaks, although their contributions
                 # to the data are still removed.
@@ -1122,7 +1122,7 @@ def extract_single(self, recursion_depth=1):
                 # near_peaks: array containing the indices of two nearest peaks on either side of border_x
                 # other_peaks: all the other peaks in new_cluster
                 # left_data, right_data: indices defining the extent of the "interpeak range" for x, etc.
-                near_peaks = np.array([], dtype=np.int)
+                near_peaks = np.array([], dtype=np.int64)
 
                 # interpeak range goes from peak to peak of next nearest peaks, although their contributions
                 # to the data are still removed.

src/diffpy/srmise/peaks/gaussian.py

@@ -464,12 +464,12 @@ def max(self, pars):
 
     guesspars = [[2.7, 0.15, 5], [3.7, 0.3, 5]]
     guess_peaks = Peaks([pf.actualize(p, "pwa") for p in guesspars])
-    cluster = ModelCluster(guess_peaks, r, y, err, None, AICc, [pf])
+    cluster = ModelCluster(guess_peaks, None, r, y, err, None, AICc, [pf])
 
     qual1 = cluster.quality()
     print(qual1.stat)
     cluster.fit()
-    yfit = cluster.calc()
+    yfit = cluster.value()
     qual2 = cluster.quality()
     print(qual2.stat)
 

src/diffpy/srmise/peaks/gaussianoverr.py

@@ -534,12 +534,12 @@ def max(self, pars):
 
     guesspars = [[2.7, 0.15, 5], [3.7, 0.3, 5]]
     guess_peaks = Peaks([pf.actualize(p, "pwa") for p in guesspars])
-    cluster = ModelCluster(guess_peaks, r, y, err, None, AICc, [pf])
+    cluster = ModelCluster(guess_peaks, None, r, y, err, None, AICc, [pf])
 
     qual1 = cluster.quality()
     print(qual1.stat)
     cluster.fit()
-    yfit = cluster.calc()
+    yfit = cluster.value()
     qual2 = cluster.quality()
     print(qual2.stat)
 

src/diffpy/srmise/peaks/terminationripples.py

@@ -316,7 +316,7 @@ def value(self, peak, r, rng=None):
             # issues is difficult to determine without detailed knowledge
             # of the underlying function.
             dr = (r[-1] - r[0]) / (len(r) - 1)
-            segments = np.ceil(dr / dr_super)
+            segments = int(np.ceil(dr / dr_super))
             dr_segmented = dr / segments
 
             rpart = r[rng]
@@ -416,12 +416,12 @@ def extend_grid(self, r, dr):
 
     guesspars = [[2.7, 0.15, 5], [3.7, 0.3, 5]]
     guess_peaks = Peaks([pf2.actualize(p, "pwa") for p in guesspars])
-    cluster = ModelCluster(guess_peaks, r, y_ripple, err, None, AICc, [pf2])
+    cluster = ModelCluster(guess_peaks, None, r, y_ripple, err, None, AICc, [pf2])
 
     qual1 = cluster.quality()
     print(qual1.stat)
     cluster.fit()
-    yfit = cluster.calc()
+    yfit = cluster.value()
     qual2 = cluster.quality()
     print(qual2.stat)