Replaces prints with logging module.

dmipy/core/modeling_framework.py

@@ -29,6 +29,7 @@
 from dipy.utils.optpkg import optional_package
 from graphviz import Digraph
 from uuid import uuid4
+import logging
 pathos, have_pathos, _ = optional_package("pathos")
 numba, have_numba, _ = optional_package("numba")
 
@@ -228,7 +229,7 @@ def parameter_initial_guess_to_parameter_vector(self, **parameters):
                 msg = '"{}" is not a valid model parameter.'.format(parameter)
                 raise ValueError(msg)
         if len(parameter_cardinality) == 0:
-            print("All model parameters set.")
+            logging.info("All model parameters set.")
         else:
             for parameter, card in parameter_cardinality.items():
                 set_parameters[parameter] = np.tile(np.nan, card)
@@ -1016,13 +1017,13 @@ def __init__(self, models, S0_tissue_responses=None, parameter_links=None):
         self.x0_parameters = {}
 
         if not have_numba:
-            msg = "We highly recommend installing numba for faster function "
-            msg += "execution and model fitting."
-            print(msg)
+            msg = "We highly recommend installing numba for faster function "\
+                  "execution and model fitting."
+            logging.info(msg)
         if not have_pathos:
             msg = "We highly recommend installing pathos to take advantage of "
             msg += "multicore processing."
-            print(msg)
+            logging.info(msg)
 
     def _check_for_NMR_and_other_models(self):
         model_types = [model._model_type for model in self.models]
@@ -1156,7 +1157,7 @@ def fit(self, acquisition_scheme, data,
             if number_of_processors is None:
                 number_of_processors = cpu_count()
             pool = pp.ProcessPool(number_of_processors)
-            print('Using parallel processing with {} workers.'.format(
+            logging.info('Using parallel processing with {} workers.'.format(
                 number_of_processors))
         else:
             fitted_parameters_lin = np.empty(
@@ -1167,12 +1168,12 @@ def fit(self, acquisition_scheme, data,
             global_brute = GlobalBruteOptimizer(
                 self, self.scheme, x0_, Ns, N_sphere_samples)
             fit_func = Brute2FineOptimizer(self, self.scheme, Ns)
-            print('Setup brute2fine optimizer in {} seconds'.format(
+            logging.info('Setup brute2fine optimizer in {} seconds'.format(
                 time() - start))
         elif solver == 'mix':
             self._check_for_tortuosity_constraint()
             fit_func = MixOptimizer(self, self.scheme, maxiter)
-            print('Setup MIX optimizer in {} seconds'.format(
+            logging.info('Setup MIX optimizer in {} seconds'.format(
                 time() - start))
         else:
             msg = "Unknown solver name {}".format(solver)
@@ -1200,15 +1201,15 @@ def fit(self, acquisition_scheme, data,
             pool.clear()
 
         fitting_time = time() - start
-        print('Fitting of {} voxels complete in {} seconds.'.format(
+        logging.info('Fitting of {} voxels complete in {} seconds.'.format(
             len(fitted_parameters_lin), fitting_time))
-        print('Average of {} seconds per voxel.'.format(
+        logging.info('Average of {} seconds per voxel.'.format(
             fitting_time / N_voxels))
 
         fitted_mt_fractions = None
         if self.S0_tissue_responses:
             # secondary fitting including S0 responses
-            print('Starting secondary multi-tissue optimization.')
+            logging.info('Starting secondary multi-tissue optimization.')
             start = time()
             mt_fractions = np.empty(
                 np.r_[N_voxels, self.N_models], dtype=float)
@@ -1220,7 +1221,7 @@ def fit(self, acquisition_scheme, data,
                 mt_fractions[idx] = fit_func(voxel_S, parameters)
             fitting_time = time() - start
             msg = 'Multi-tissue fitting of {} voxels complete in {} seconds.'
-            print(msg.format(len(mt_fractions), fitting_time))
+            logging.info(msg.format(len(mt_fractions), fitting_time))
             fitted_mt_fractions = np.zeros(np.r_[mask.shape, self.N_models])
             fitted_mt_fractions[mask_pos] = mt_fractions
 
@@ -1396,11 +1397,11 @@ def __init__(self, models, S0_tissue_responses=None, parameter_links=None):
         if not have_numba:
             msg = "We highly recommend installing numba for faster function "
             msg += "execution and model fitting."
-            print(msg)
+            logging.info(msg)
         if not have_pathos:
             msg = "We highly recommend installing pathos to take advantage of "
             msg += "multicore processing."
-            print(msg)
+            logging.info(msg)
 
     def _check_for_NMR_models(self):
         for model in self.models:
@@ -1551,7 +1552,7 @@ def fit(self, acquisition_scheme, data,
             if number_of_processors is None:
                 number_of_processors = cpu_count()
             pool = pp.ProcessPool(number_of_processors)
-            print('Using parallel processing with {} workers.'.format(
+            logging.info('Using parallel processing with {} workers.'.format(
                 number_of_processors))
         else:
             fitted_parameters_lin = np.empty(
@@ -1569,12 +1570,12 @@ def fit(self, acquisition_scheme, data,
                 self, self.scheme,
                 x0_, Ns, N_sphere_samples)
             fit_func = Brute2FineOptimizer(self, self.scheme, Ns)
-            print('Setup brute2fine optimizer in {} seconds'.format(
+            logging.info('Setup brute2fine optimizer in {} seconds'.format(
                 time() - start))
         elif solver == 'mix':
             self._check_for_tortuosity_constraint()
             fit_func = MixOptimizer(self, self.scheme, maxiter)
-            print('Setup MIX optimizer in {} seconds'.format(
+            logging.info('Setup MIX optimizer in {} seconds'.format(
                 time() - start))
         else:
             msg = "Unknown solver name {}".format(solver)
@@ -1602,15 +1603,15 @@ def fit(self, acquisition_scheme, data,
             pool.clear()
 
         fitting_time = time() - start
-        print('Fitting of {} voxels complete in {} seconds.'.format(
+        logging.info('Fitting of {} voxels complete in {} seconds.'.format(
             len(fitted_parameters_lin), fitting_time))
-        print('Average of {} seconds per voxel.'.format(
+        logging.info('Average of {} seconds per voxel.'.format(
             fitting_time / N_voxels))
 
         fitted_mt_fractions = None
         if self.S0_tissue_responses:
             # secondary fitting including S0 responses
-            print('Starting secondary multi-tissue optimization.')
+            logging.info('Starting secondary multi-tissue optimization.')
             start = time()
             mt_fractions = np.empty(
                 np.r_[N_voxels, self.N_models], dtype=float)
@@ -1622,7 +1623,7 @@ def fit(self, acquisition_scheme, data,
                 mt_fractions[idx] = fit_func(voxel_S, parameters)
             fitting_time = time() - start
             msg = 'Multi-tissue fitting of {} voxels complete in {} seconds.'
-            print(msg.format(len(mt_fractions), fitting_time))
+            logging.info(msg.format(len(mt_fractions), fitting_time))
             fitted_mt_fractions = np.zeros(np.r_[mask.shape, self.N_models])
             fitted_mt_fractions[mask_pos] = mt_fractions
 
@@ -1790,11 +1791,11 @@ def __init__(self, models, S0_tissue_responses=None, sh_order=8):
         if not have_numba:
             msg = "We highly recommend installing numba for faster function "
             msg += "execution and model fitting."
-            print(msg)
+            logging.info(msg)
         if not have_pathos:
             msg = "We highly recommend installing pathos to take advantage of "
             msg += "multicore processing."
-            print(msg)
+            logging.info(msg)
 
     def _check_for_dispersed_or_NMR_models(self):
         for model in self.models:
@@ -2012,10 +2013,10 @@ def fit(self, acquisition_scheme, data, mask=None, solver='csd',
                     msg += ' optimizer because it does not improve fitting '
                     msg += 'speed.'
                     if verbose:
-                        print(msg)
+                        logging.info(msg)
                     use_parallel_processing = False
                 if verbose:
-                    print(
+                    logging.info(
                         'Setup Tournier07 FOD optimizer in {} seconds'.format(
                             time() - start))
             else:
@@ -2024,8 +2025,8 @@ def fit(self, acquisition_scheme, data, mask=None, solver='csd',
                     unity_constraint=self.unity_constraint,
                     lambda_lb=lambda_lb)
                 if verbose:
-                    print('Setup CVXPY FOD optimizer in {} seconds'.format(
-                        time() - start))
+                    msg = 'Setup CVXPY FOD optimizer in {} seconds'
+                    logging.info(msg.format(time() - start))
         elif solver == 'csd_tournier07':
             fit_func = CsdTournierOptimizer(
                 acquisition_scheme, self, x0_, self.sh_order,
@@ -2034,18 +2035,18 @@ def fit(self, acquisition_scheme, data, mask=None, solver='csd',
                 msg = 'Parallel processing turned off for tournier07 optimizer'
                 msg += ' because it does not improve fitting speed.'
                 if verbose:
-                    print(msg)
+                    logging.info(msg)
                 use_parallel_processing = False
             if verbose:
-                print('Setup Tournier07 FOD optimizer in {} seconds'.format(
-                    time() - start))
+                msg = 'Setup Tournier07 FOD optimizer in {} seconds'
+                logging.info(msg.format(time() - start))
         elif solver == 'csd_cvxpy':
             fit_func = CsdCvxpyOptimizer(
                 acquisition_scheme, self, x0_, self.sh_order,
                 unity_constraint=self.unity_constraint, lambda_lb=lambda_lb)
             if verbose:
-                print('Setup CVXPY FOD optimizer in {} seconds'.format(
-                    time() - start))
+                msg = 'Setup CVXPY FOD optimizer in {} seconds'
+                logging.info(msg.format(time() - start))
         else:
             msg = "Unknown solver name {}".format(solver)
             raise ValueError(msg)
@@ -2061,8 +2062,8 @@ def fit(self, acquisition_scheme, data, mask=None, solver='csd',
                 number_of_processors = cpu_count()
             pool = pp.ProcessPool(number_of_processors)
             if verbose:
-                print('Using parallel processing with {} workers.'.format(
-                    number_of_processors))
+                msg = 'Using parallel processing with {} workers.'
+                logging.info(msg.format(number_of_processors))
         else:
             fitted_parameters_lin = np.empty(
                 np.r_[N_voxels, N_parameters], dtype=float)
@@ -2089,9 +2090,9 @@ def fit(self, acquisition_scheme, data, mask=None, solver='csd',
 
         fitting_time = time() - start
         if verbose:
-            print('Fitting of {} voxels complete in {} seconds.'.format(
+            logging.info('Fitting of {} voxels complete in {} seconds.'.format(
                 len(fitted_parameters_lin), fitting_time))
-            print('Average of {} seconds per voxel.'.format(
+            logging.info('Average of {} seconds per voxel.'.format(
                 fitting_time / N_voxels))
         fitted_parameters = np.zeros_like(x0_, dtype=float)
         fitted_parameters[mask_pos] = fitted_parameters_lin

dmipy/custom_optimizers/intra_voxel_incoherent_motion.py

@@ -2,6 +2,7 @@
 from dmipy.core.modeling_framework import MultiCompartmentModel
 import numpy as np
 from time import time
+import logging
 
 
 def ivim_Dstar_fixed(acquisition_scheme, data, mask=None, Dstar_value=7e-9,
@@ -60,7 +61,7 @@ def ivim_Dstar_fixed(acquisition_scheme, data, mask=None, Dstar_value=7e-9,
     if fit_args is None:
         fit_args = {}
 
-    print('Starting IVIM Dstar-fixed algorithm.')
+    logging.info('Starting IVIM Dstar-fixed algorithm.')
     ivim_mod = MultiCompartmentModel([G1Ball(), G1Ball()])
     ivim_mod.set_fixed_parameter(
         'G1Ball_2_lambda_iso', Dstar_value)  # following [2]
@@ -76,5 +77,5 @@ def ivim_Dstar_fixed(acquisition_scheme, data, mask=None, Dstar_value=7e-9,
     N_voxels = np.sum(ivim_fit.mask)
     msg = 'IVIM Dstar-fixed optimization of {0:d} voxels'.format(N_voxels)
     msg += ' complete in {0:.3f} seconds'.format(computation_time)
-    print(msg)
+    logging.info(msg)
     return ivim_fit

dmipy/custom_optimizers/single_shell_three_tissue_csd.py

@@ -4,6 +4,7 @@
     MultiCompartmentSphericalHarmonicsModel)
 from dmipy.tissue_response.three_tissue_response import (
     three_tissue_response_dhollander16)
+import logging
 
 
 def single_shell_three_tissue_csd(
@@ -116,7 +117,7 @@ def single_shell_three_tissue_csd(
             mt_csd_fits.append(mt_csd_fit)
         computation_time = time.time() - start
         if ss3t_verbose:
-            print('finish it {} of {} in {} seconds'.format(
+            logging.info('finish it {} of {} in {} seconds'.format(
                 it + 1, N_iterations, int(computation_time)))
     if return_all_csd_fits:
         return mt_csd_fits

dmipy/data/saved_data.py

@@ -7,6 +7,7 @@
 import matplotlib.pyplot as plt
 import zipfile
 from . import saved_acquisition_schemes
+import logging
 
 try:
     from urllib2 import urlopen
@@ -76,7 +77,7 @@ def wu_minn_hcp_coronal_slice():
     msg += "NIH Institutes and Centers that support the NIH Blueprint for "
     msg += "Neuroscience Research; and by the McDonnell Center for Systems "
     msg += "Neuroscience at Washington University."
-    print(msg)
+    logging.info(msg)
 
     return scheme, data
 
@@ -88,7 +89,7 @@ def duval_cat_spinal_cord_2d():
     msg += "segmentation', ISMRM 2016. Reference at "
     msg += "Cohen-Adad et al. White Matter Microscopy Database."
     msg += " http://doi.org/10.17605/OSF.IO/YP4QG"
-    print(msg)
+    logging.info(msg)
 
     data_folder = join(DATA_PATH, "tanguy_cat_spinal_cord")
 
@@ -130,7 +131,7 @@ def duval_cat_spinal_cord_3d():
     msg += "segmentation', ISMRM 2016. Reference at "
     msg += "Cohen-Adad et al. White Matter Microscopy Database."
     msg += " http://doi.org/10.17605/OSF.IO/YP4QG"
-    print(msg)
+    logging.info(msg)
 
     data_folder = join(DATA_PATH, "tanguy_cat_spinal_cord")
 

dmipy/distributions/distribute_models.py

@@ -6,6 +6,7 @@
 from ..core.signal_model_properties import AnisotropicSignalModelProperties
 import copy
 import numpy as np
+import logging
 
 __all__ = [
     'DistributedModel',
@@ -301,8 +302,8 @@ def set_tortuous_parameter(self, lambda_perp,
             try:
                 self.parameter_ranges[param]
             except KeyError:
-                print("{} does not exist or has already been fixed.").format(
-                    param)
+                msg = "{} does not exist or has already been fixed."
+                logging.error(msg.format(param))
                 return None
 
         model, name = self._parameter_map[lambda_perp]
@@ -337,9 +338,10 @@ def set_equal_parameter(self, parameter_name_in, parameter_name_out):
             try:
                 self.parameter_ranges[param]
             except KeyError:
-                print("{} does not exist or has already been fixed.").format(
-                    param)
+                msg = "{} does not exist or has already been fixed."
+                logging.error(msg.format(param))
                 return None
+
         model, name = self._parameter_map[parameter_name_out]
         self.parameter_links.append([model, name, parameter_equality, [
             self._parameter_map[parameter_name_in]]])

dmipy/hcp_interface/downloader_aws.py

@@ -4,6 +4,7 @@
 import os
 import nibabel as nib
 import numpy as np
+import logging
 
 
 DATA_PATH = pkg_resources.resource_filename(
@@ -85,7 +86,7 @@ def download_subject(self, subject_ID):
         if not os.path.exists(hcp_data_path):
             os.makedirs(hcp_data_path)
 
-        print('Downloading data to {}'.format(hcp_data_path))
+        logging.info('Downloading data to {}'.format(hcp_data_path))
 
         counter = 0
         for key in self.s3_bucket.list("HCP_1200"):
@@ -101,7 +102,7 @@ def download_subject(self, subject_ID):
                     'data' in path.parts[-1] or
                     'nodif' in path.parts[-1]
                 ):
-                    print('Downloading {}'.format(path.parts[-1]))
+                    logging.info('Downloading {}'.format(path.parts[-1]))
                     filepath = os.path.join(hcp_data_path, path.parts[-1])
                     with open(filepath, 'wb') as f:
                         key.get_contents_to_file(f)
@@ -112,7 +113,7 @@ def download_subject(self, subject_ID):
     def prepare_example_slice(self, subject_ID):
         "Prepares a coronal slice for the dmipy example notebooks."
         msg = "Preparing coronal slice for dmipy examples"
-        print(msg)
+        logging.info(msg)
 
         folder_name = "hcp_example_slice"
         example_directory = os.path.join(self.hcp_directory, folder_name)
@@ -136,4 +137,4 @@ def prepare_example_slice(self, subject_ID):
 
         nib.save(nib.Nifti1Image(data_slice, affine), os.path.join(
             example_directory, 'coronal_slice.nii.gz'))
-        print('Done')
+        logging.info('Done')