ajc: rename ajcs to ajc

NEWS.md

@@ -1,7 +1,12 @@
 current
 ====
-* New augmented jump chain module
+* ajc.py: New augmented jump chain module
   * Sparse implementation for the computation of the transition kernel, koopman operators and space-time-committors
+* schur.py
+  * allow sensivity and iteration adjustment for the KrylovSchur solver
+  * change tolerance to absolute value
+* pcca.py
+  * fix memory allocation for sparse systems
 
 0.2.1
 =====

README.md

@@ -14,7 +14,7 @@ This Python library implements a suite of tools used and/or developed in the [Co
 
 ## Contents
 *  `pcca`: An implementation of (generalized) PCCA⁺ using the Schur decomposition
-*  `ajcs`: A sparse implementation of the augmented jump chain
+*  `ajc`: A sparse implementation of the augmented jump chain
 *  `diffusionmaps`: Diffusionmaps with sparse support and out of sample extensions
 *  `galerkin`: Trajectory based estimation of the transfer operator using a Galerkin projection onto Gaussian RBFs
 *  `gillespie`: Trajectory simulation from a generator

examples/examples.ipynb

@@ -308,7 +308,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "ajc = cmdtools.estimation.ajcs.AJCS(qs, dts) # construct the augmented jump chain"
+    "ajc = cmdtools.estimation.ajc.AJC(qs, dts) # construct the augmented jump chain"
    ]
   },
   {

src/cmdtools/__init__.py

@@ -1,3 +1,3 @@
 from .analysis import pcca, schur, optimization
-from .estimation import diffusionmaps, galerkin, gillespie, newton_generator, picking_algorithm, sqra, voronoi, ajcs
+from .estimation import diffusionmaps, galerkin, gillespie, newton_generator, picking_algorithm, sqra, voronoi, ajc
 from .systems import diffusion

src/cmdtools/estimation/ajcs.py → src/cmdtools/estimation/ajc.py

@@ -12,7 +12,7 @@
 from scipy.optimize import minimize
 
 
-class AJCS():
+class AJC():
     """ sparse implementation of the Galerkin discretization to the AJC as in the preprint
     uses the """
 

tests/test_ajc.py

@@ -1,13 +1,13 @@
 from cmdtools.systems import diffusion
-from cmdtools.estimation import ajcs
+from cmdtools.estimation import ajc
 import numpy as np
 from scipy.linalg import expm
 
 
 def test_ajc(n=20, tol=1e-2):
     Q = diffusion.DoubleWell().Q
     ts = np.repeat(1/n, n)
-    a = ajcs.AJCS(np.repeat(Q, n), ts)
+    a = ajc.AJC(np.repeat(Q, n), ts)
 
     assert np.isclose(expm(Q).todense(), a.koopman(), atol=tol).all()
 
@@ -19,7 +19,7 @@ def test_ajc(n=20, tol=1e-2):
 
 def test_limit(tol=1e-5):
     Q = diffusion.DoubleWell().Q
-    a0 = ajcs.AJCS([Q, Q, Q], [0,1,0])
-    a1 = ajcs.AJCS([Q, Q, Q], [1e-10, 1, 1e-10])
+    a0 = ajc.AJC([Q, Q, Q], [0,1,0])
+    a1 = ajc.AJC([Q, Q, Q], [1e-10, 1, 1e-10])
 
     assert np.isclose(a1.koopman(), a0.koopman(), atol=tol).all()