Merge pull request #56 from dkazanc/tests

unittests added
dkazanc · Jun 25, 2020 · 23805fc · 23805fc
2 parents f098813 + bf15510
commit 23805fc
Show file tree

Hide file tree

Showing 10 changed files with 126 additions and 76 deletions.
diff --git a/.travis.yml b/.travis.yml
@@ -1,13 +1,6 @@
 language: python
 
 matrix:
-  include:
-    - python: 2.7
-      env: NUMPY=1.16      
-
-    - python: 2.7
-      env: NUMPY=1.15 
-
     - python: 3.5
       env: NUMPY=1.15
 
@@ -16,7 +9,13 @@ matrix:
 
     - python: 3.6
       env: NUMPY=1.16
-
+
+    - python: 3.7
+      env: NUMPY=1.15
+
+    - python: 3.7
+      env: NUMPY=1.16      
+
 os:
   - linux
 
@@ -41,10 +40,10 @@ install:
   - export VERSION=`date +%Y.%m`
   - conda build conda-recipe --numpy=$NUMPY --python=$TRAVIS_PYTHON_VERSION
   - conda install --channel /home/travis/miniconda/envs/test-environment/conda-bld/ tomobar --offline --override-channels
-
+  
 after_success:
   - chmod +x src/Python/conda-recipe/conda_upload.sh
   - test $TRAVIS_BRANCH = "master" && bash conda-recipe/conda_upload.sh
 
 script:
-  - python tests/test.py
+  - python test/test_tomobarCPU_DIR.py
diff --git a/Demos/Python/DemoFISTA_artifacts2D.py b/Demos/Python/DemoFISTA_artifacts2D.py
@@ -117,7 +117,7 @@
 plt.title('Misaligned noisy FBP Reconstruction')
 plt.show()
 
-plt.figure() 
+plt.figure()
 plt.imshow(abs(FBPrec_ideal-FBPrec_error), vmin=0, vmax=2, cmap="gray")
 plt.colorbar(ticks=[0, 0.5, 2], orientation='vertical')
 plt.title('FBP reconstruction differences')
@@ -133,7 +133,7 @@
 
 FBPrec_misalign = Rectools.FBP(sino_misalign) # reconstruction with misalignment
 
-plt.figure() 
+plt.figure()
 plt.imshow(FBPrec_misalign, vmin=0, vmax=1, cmap="gray")
 plt.title('FBP reconstruction of misaligned data using known exact shifts')
 #%%
@@ -153,7 +153,7 @@
 # data dictionary
 _data_ = {'projection_norm_data' : sino_artifacts,
           'projection_raw_data' : sino_artifacts_raw/np.max(sino_artifacts_raw),
-           'OS_number' : 10} 
+           'OS_number' : 10}
 lc = RectoolsIR.powermethod(_data_) # calculate Lipschitz constant (run once to initialise)
 
 _algorithm_ = {'iterations' : 30,
@@ -169,7 +169,7 @@
 print("Run FISTA reconstrucion algorithm with regularisation...")
 RecFISTA_LS_reg = RectoolsIR.FISTA(_data_, _algorithm_, _regularisation_)
 
-# adding Huber data fidelity threshold 
+# adding Huber data fidelity threshold
 _data_.update({'huber_threshold' : 7.0})
 print(" Run FISTA reconstrucion algorithm with regularisation and Huber data...")
 RecFISTA_Huber_reg = RectoolsIR.FISTA(_data_, _algorithm_, _regularisation_)
@@ -196,7 +196,7 @@
 plt.title('FISTA-HuberRing-TV reconstruction')
 plt.show()
 
-# calculate errors 
+# calculate errors
 Qtools = QualityTools(phantom_2D[indicesROI], RecFISTA_LS_reg[indicesROI])
 RMSE_FISTA_LS_TV = Qtools.rmse()
 Qtools = QualityTools(phantom_2D[indicesROI], RecFISTA_Huber_reg[indicesROI])
@@ -256,7 +256,7 @@
 plt.title('FISTA-SWLS-Huber-TV reconstruction')
 plt.show()
 
-# calculate errors 
+# calculate errors
 Qtools = QualityTools(phantom_2D[indicesROI], RecFISTA_SWLS[indicesROI])
 RMSE_FISTA_SWLS_TV = Qtools.rmse()
 Qtools = QualityTools(phantom_2D[indicesROI], RecFISTA_SWLS_Huber[indicesROI])
@@ -279,7 +279,7 @@
                'mask_diameter' : 0.9,
                'lipschitz_const' : lc}
 
-# Run FISTA reconstrucion algorithm with regularisation 
+# Run FISTA reconstrucion algorithm with regularisation
 RecFISTA_LS_GH_reg = RectoolsIR.FISTA(_data_, _algorithm_, _regularisation_)
 
 plt.figure()
@@ -288,7 +288,7 @@
 plt.title('FISTA-OS-GH-TV reconstruction')
 plt.show()
 
-# calculate errors 
+# calculate errors
 Qtools = QualityTools(phantom_2D[indicesROI], RecFISTA_LS_GH_reg[indicesROI])
 RMSE_FISTA_LS_GH_TV = Qtools.rmse()
 print("RMSE for FISTA-PWLS-GH-TV reconstruction is {}".format(RMSE_FISTA_LS_GH_TV))
@@ -312,7 +312,7 @@
                     'iterations' : 120,
                     'device_regulariser': 'gpu'}
 
-# Run FISTA reconstrucion algorithm with regularisation 
+# Run FISTA reconstrucion algorithm with regularisation
 RecFISTA_LS_stud_reg = RectoolsIR.FISTA(_data_, _algorithm_, _regularisation_)
 
 plt.figure()
@@ -321,8 +321,8 @@
 plt.title('FISTA-OS-Stidentst-TV reconstruction')
 plt.show()
 
-# calculate errors 
+# calculate errors
 Qtools = QualityTools(phantom_2D[indicesROI], RecFISTA_LS_stud_reg[indicesROI])
 RMSE_FISTA_LS_studentst_TV = Qtools.rmse()
 print("RMSE for FISTA-PWLS-Studentst-TV reconstruction is {}".format(RMSE_FISTA_LS_studentst_TV))
-#%%
+#%%
diff --git a/Demos/Python/DemoFISTA_artifacts3D.py b/Demos/Python/DemoFISTA_artifacts3D.py
@@ -50,7 +50,7 @@
 
 # Projection geometry related parameters:
 Horiz_det = int(N_size) # detector column count (horizontal)
-Vert_det = N_size+20 # detector row count (vertical) (no reason for it to be > N)
+Vert_det = N_size # detector row count (vertical) (no reason for it to be > N)
 angles_num = int(0.5*np.pi*N_size); # angles number
 angles = np.linspace(0.0,179.9,angles_num,dtype='float32') # in degrees
 angles_rad = angles*(np.pi/180.0)

diff --git a/conda-recipe/build.sh b/conda-recipe/build.sh
@@ -1,4 +1,5 @@
-set -xe 
+set -xe
+cp -rv "$RECIPE_DIR/../test" "$SRC_DIR/"
 
 cd $SRC_DIR
 

diff --git a/conda-recipe/meta.yaml b/conda-recipe/meta.yaml
@@ -1,22 +1,29 @@
 package:
   name: tomobar
   version: {{ environ['VERSION'] }}
-  
+
 build:
   preserve_egg_dir: False
   number: 0
   script_env:
     - VERSION
 
+test:
+  source_files:
+    - ./test/
+  commands:
+    - python -c "import os; print (os.getcwd())"
+    - python -m unittest discover -s test
+
 requirements:
   build:
     - python
     - numpy
     - setuptools
     - cython
     - cmake
-
   run:
+    - scipy
     - python
     - numpy
     - libgcc-ng # [unix]

diff --git a/src/Python/tomobar/methodsIR.py b/src/Python/tomobar/methodsIR.py
@@ -22,17 +22,28 @@
 
 import numpy as np
 from numpy import linalg as LA
-import scipy.sparse.linalg
 
 try:
-    from tomobar.supp.addmodules import RING_WEIGHTS
-except:    
-    print('____! RING_WEIGHTS C-module failed on import !____')
+    from ccpi.filters.regularisers import ROF_TV,FGP_TV,PD_TV,SB_TV,LLT_ROF,TGV,NDF,Diff4th,NLTV
+except:
+    print('____! CCPi-regularisation package is missing, please install !____')
 
 try:
-    from ccpi.filters.regularisers import ROF_TV,FGP_TV,PD_TV,SB_TV,LLT_ROF,TGV,NDF,Diff4th,NLTV
+    import astra
+except:
+    print('____! Astra-toolbox package is missing, please install !____')
+
+try:
+    import scipy.sparse.linalg
+except:
+    print('____! Scipy toolbox package is missing, please install !____')
+
+try:
+    from tomobar.supp.addmodules import RING_WEIGHTS
 except:
-    print('____! CCPi regularisation package is missing, please install !____')
+    print('____! RING_WEIGHTS C-module failed on import !____')
+
+
 
 def smooth(y, box_pts):
     # a function to smooth 1D signal
@@ -264,7 +275,7 @@ class RecToolsIR:
             --huber_threshold # threshold for Huber function to apply to data model (supress outliers)
             --studentst_threshold # threshold for Students't function to apply to data model (supress outliers)
             --ring_weights_threshold # threshold to produce additional weights to supress ring artifacts
-            --ring_huber_power # defines the strength of Huber penalty to supress artifacts 1 = Huber, > 1 more penalising  
+            --ring_huber_power # defines the strength of Huber penalty to supress artifacts 1 = Huber, > 1 more penalising
             --ring_tuple_halfsizes # a tuple for half window sizes as [detector, angles, num of projections]
             --ringGH_lambda # a parameter for Group-Huber data model to supress full rings of the same intensity
             --ringGH_accelerate # Group-Huber data model acceleration factor (use carefully to avoid divergence, 50 default)
@@ -347,7 +358,7 @@ def __init__(self,
             from tomobar.supp.astraOP import AstraTools3D
             self.Atools = AstraTools3D(self.DetectorsDimH, self.DetectorsDimV, self.AnglesVec, self.CenterRotOffset, self.ObjSize) # initiate 3D ASTRA class object
         return None
-            
+
 
     def SIRT(self, _data_, _algorithm_):
         ######################################################################
@@ -372,7 +383,7 @@ def CGLS(self, _data_, _algorithm_):
         if (self.geom == '3D'):
             CGLS_rec = self.Atools.cgls3D(_data_['projection_norm_data'], _algorithm_['iterations'])
         return CGLS_rec
-    
+
     def powermethod(self, _data_):
         # power iteration algorithm to  calculate the eigenvalue of the operator (projection matrix)
         # projection_raw_data is required for PWLS fidelity (self.datafidelity = PWLS), otherwise will be ignored
@@ -389,8 +400,8 @@ def powermethod(self, _data_):
                 self.AtoolsOS = AstraToolsOS3D(self.DetectorsDimH, self.DetectorsDimV, self.AnglesVec, self.CenterRotOffset, self.ObjSize, _data_['OS_number']) # initiate 3D ASTRA class OS object
         niter = 15 # number of power method iterations
         s = 1.0
-        
-        # classical approach 
+
+        # classical approach
         if (self.geom == '2D'):
             x1 = np.float32(np.random.randn(self.Atools.ObjSize,self.Atools.ObjSize))
         else:
@@ -489,9 +500,9 @@ def FISTA(self, _data_, _algorithm_, _regularisation_):
                     r[:,0] = r_x[:,0] - np.multiply(L_const_inv,vec)
                 else:
                     r = r_x - np.multiply(L_const_inv,vec)
-            
+
             if ((_data_['OS_number'] != 1) and (_data_['ring_weights_threshold'] is not None) and (iter > 0)):
-                # Ordered subset approach for a better ring model 
+                # Ordered subset approach for a better ring model
                 res_full = self.Atools.forwproj(X_t) - _data_['projection_norm_data']
                 rings_weights = RING_WEIGHTS(res_full, _data_['ring_tuple_halfsizes'][0], _data_['ring_tuple_halfsizes'][1], _data_['ring_tuple_halfsizes'][2])
                 ring_function_weight = np.ones(np.shape(res_full))
@@ -703,7 +714,7 @@ def ADMM_Atb(b):
             # update u variable
             u = u + (x_hat - z)
             if (_algorithm_['verbose'] == 'on'):
-                if (np.mod(iter,(round)(_algorithm_['iterations']/5)+1) == 0):                    
+                if (np.mod(iter,(round)(_algorithm_['iterations']/5)+1) == 0):
                     print('ADMM iteration (',iter+1,') using', _regularisation_['method'], 'regularisation for (',(int)(info_vec[0]),') iterations')
             if (iter == _algorithm_['iterations']-1):
                 print('ADMM stopped at iteration (', iter+1, ')')