Merge branch 'diffpy:main' into readme

Author: stevenhua0320

.github/workflows/main.yml

@@ -44,7 +44,7 @@ jobs:
           conda activate test
           conda install --file requirements/run.txt
           conda install --file requirements/test.txt
-          pip install .
+          pip install . --no-deps
 
       - name: Validate diffpy.srmise
         shell: bash -l {0}

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "diffpy.srmise"
-dynamic=['version']
+dynamic=['version', 'dependencies']
 authors = [
   { name="Simon J.L. Billinge group", email="[email protected]" },
         {name="Luke Granlund", email="[email protected]"},
@@ -51,6 +51,9 @@ include = ["*"]  # package names should match these glob patterns (["*"] by defa
 exclude = ["diffpy.srmise.tests*"]  # exclude packages matching these glob patterns (empty by default)
 namespaces = false  # to disable scanning PEP 420 namespaces (true by default)
 
+[tool.setuptools.dynamic]
+dependencies = {file = ["requirements/run.txt"]}
+
 [tool.black]
 line-length = 115
 include = '\.pyi?$'

src/diffpy/srmise/baselines/polynomial.py

@@ -105,7 +105,7 @@ def estimate_parameters(self, r, y):
                 import numpy.linalg as la
 
                 a = np.array([r[cut_idx]]).T
-                slope = la.lstsq(a, y[cut_idx])[0][0]
+                slope = la.lstsq(a, y[cut_idx], rcond=-1)[0][0]
                 return np.array([slope, 0.0])
             except Exception as e:
                 emsg = "Error during estimation -- " + str(e)
@@ -256,4 +256,5 @@ def getmodule(self):
     y = -r + 10 * np.exp(-((r - 5) ** 2)) + np.random.rand(len(r))
     est = f.estimate_parameters(r, y)
     print("Actual baseline: ", np.array([-1, 0.0]))
+    # TODO: Make test est baseline in ways of tolerance function
     print("Estimated baseline: ", est)

src/diffpy/srmise/dataclusters.py

@@ -40,15 +40,15 @@ class DataClusters:
       The array of r values.
     y : sequence of y values
       The array of PDF values, G(r)
-    res : int
-      The clustering resolution, i.e., the number of points another point has to
+    res : float
+      The clustering resolution, i.e., the number of distance another point has to
       be away from the center of an existing cluster to before a new cluster is
-      formed.  A value of zero allows every point to be cluster.
+      formed.  A value of zero allows every point to be a cluster.
     data_order : array
       The array of x, y indices ordered by decreasing y
-    clusters :
+    clusters : ndarray
       The array of cluster ranges
-    current_idx - int
+    current_idx : int
       The index of data_order currently considered
     """
 
@@ -61,8 +61,8 @@ def __init__(self, x, y, res):
           The array of r values.
         y : sequence of y values
           The array of PDF values, G(r)
-        res : int
-          The clustering resolution, i.e., the number of points another point has to
+        res : float
+          The clustering resolution, i.e., the distance another point has to
           be away from the center of an existing cluster to before a new cluster is
           formed.  A value of zero allows every point to be cluster.
         """
@@ -143,8 +143,8 @@ def _setdata(self, x, y, res):
           The array of r values.
         y : sequence of y values
           The array of PDF values, G(r)
-        res : int
-          The clustering resolution, i.e., the number of points another point has to
+        res : float
+          The clustering resolution, i.e., the distance another point has to
           be away from the center of an existing cluster to before a new cluster is
           formed.  A value of zero allows every point to be cluster.
         """
@@ -230,13 +230,18 @@ def find_nearest_cluster2(self, x):
         """Return [cluster index, distance] for cluster nearest to x.
 
         Parameters
-        x - Coordinate of point of interest
+        ----------
+        x : ndarray
+            Coordinate of point of interest
 
         The distance is positive/negative if the point is right/left of the
         nearest cluster.  If the point is within an existing cluster then
         distance = 0.
 
-        Return None if no clusters exists.
+        Returns
+        -------
+        array-like
+            The index of the nearest cluster, and the distance for cluster nearest to x. None if no cluster
         """
         if self.status == self.INIT:
             raise Exception("Cannot cluster next point while status is INIT.")
@@ -259,9 +264,14 @@ def find_nearest_cluster(self, idx):
         distance = 0.
 
         Parameters
-        idx - index of point in self.x of interest.
+        ----------
+        idx : array-like
+            index of point in self.x of interest.
 
-        Return None if no clusters exist.
+        Returns
+        -------
+        array-like
+            The array of cluster index and the distacne to the nearest cluster. None if no clusters exist.
         """
         if self.status == self.INIT:
             raise Exception("Cannot cluster next point while status is INIT.")
@@ -302,7 +312,14 @@ def cluster_is_full(self, cluster_idx):
         boundaries.
 
         Parameters
-        cluster_idx - The index of the cluster to test
+        ----------
+        cluster_idx : array-like
+            The index of the cluster to test
+
+        Returns
+        -------
+        bools
+            True if the cluster is full, False otherwise
         """
         if cluster_idx > 0:
             low = self.clusters[cluster_idx - 1, 1] + 1
@@ -321,7 +338,13 @@ def combine_clusters(self, combine):
         unclustered points between them.
 
         Parameters
-        combine - [[leftmost_idx1, ..., rightmost_idx1], ...]
+        ----------
+        combine : ndarray
+            [[leftmost_idx1, ..., rightmost_idx1], ...]
+
+        Returns
+        -------
+        None
         """
         # Ensure that the same clusters aren't combined multiple times.
         combine_flat = np.array(combine).ravel()