Remove n_procs for nodes that do not use multiple cores

xcp_d/workflows/anatomical/plotting.py

@@ -68,7 +68,6 @@ def init_brainsprite_figures_wf(t1w_available, t2w_available, name="brainsprite_
     workflow = Workflow(name=name)
 
     output_dir = config.execution.xcp_d_dir
-    omp_nthreads = config.nipype.omp_nthreads
 
     inputnode = pe.Node(
         niu.IdentityInterface(
@@ -108,7 +107,6 @@ def init_brainsprite_figures_wf(t1w_available, t2w_available, name="brainsprite_
             ),
             name=f"get_number_of_frames_{image_type}",
             mem_gb=config.DEFAULT_MEMORY_MIN_GB,
-            omp_nthreads=omp_nthreads,
         )
         workflow.connect([
             (inputnode, get_number_of_frames, [(inputnode_anat_name, "anat_file")]),
@@ -132,7 +130,6 @@ def init_brainsprite_figures_wf(t1w_available, t2w_available, name="brainsprite_
             name=f"modify_brainsprite_template_scene_{image_type}",
             iterfield=["slice_number"],
             mem_gb=config.DEFAULT_MEMORY_MIN_GB,
-            omp_nthreads=omp_nthreads,
         )
         modify_brainsprite_template_scene.inputs.scene_template = brainsprite_scene_template
         workflow.connect([
@@ -157,7 +154,6 @@ def init_brainsprite_figures_wf(t1w_available, t2w_available, name="brainsprite_
             name=f"create_framewise_pngs_{image_type}",
             iterfield=["scene_file"],
             mem_gb=1,
-            omp_nthreads=omp_nthreads,
         )
         workflow.connect([
             (modify_brainsprite_template_scene, create_framewise_pngs, [
@@ -174,7 +170,6 @@ def init_brainsprite_figures_wf(t1w_available, t2w_available, name="brainsprite_
             ),
             name=f"make_mosaic_{image_type}",
             mem_gb=1,
-            omp_nthreads=omp_nthreads,
         )
 
         workflow.connect([(create_framewise_pngs, make_mosaic_node, [("out_file", "png_files")])])
@@ -211,7 +206,6 @@ def init_brainsprite_figures_wf(t1w_available, t2w_available, name="brainsprite_
             ),
             name=f"modify_pngs_template_scene_{image_type}",
             mem_gb=config.DEFAULT_MEMORY_MIN_GB,
-            omp_nthreads=omp_nthreads,
         )
         modify_pngs_template_scene.inputs.scene_template = pngs_scene_template
         workflow.connect([
@@ -238,7 +232,6 @@ def init_brainsprite_figures_wf(t1w_available, t2w_available, name="brainsprite_
             name=f"create_scenewise_pngs_{image_type}",
             iterfield=["scene_name_or_number"],
             mem_gb=1,
-            omp_nthreads=omp_nthreads,
         )
         workflow.connect([
             (modify_pngs_template_scene, create_scenewise_pngs, [("out_file", "scene_file")]),

xcp_d/workflows/anatomical/surface.py

@@ -387,7 +387,6 @@ def init_warp_surfaces_to_template_wf(
     # First, we create the Connectome WorkBench-compatible transform files.
     update_xfm_wf = init_ants_xfm_to_fsl_wf(
         mem_gb=1,
-        omp_nthreads=omp_nthreads,
         name="update_xfm_wf",
     )
     workflow.connect([
@@ -510,7 +509,6 @@ def init_generate_hcp_surfaces_wf(name="generate_hcp_surfaces_wf"):
     workflow = Workflow(name=name)
 
     output_dir = config.execution.xcp_d_dir
-    omp_nthreads = config.nipype.omp_nthreads
 
     inputnode = pe.Node(
         niu.IdentityInterface(
@@ -532,7 +530,6 @@ def init_generate_hcp_surfaces_wf(name="generate_hcp_surfaces_wf"):
         SurfaceAverage(),
         name="generate_midthickness",
         mem_gb=2,
-        n_procs=omp_nthreads,
     )
     workflow.connect([
         (inputnode, generate_midthickness, [
@@ -565,7 +562,6 @@ def init_generate_hcp_surfaces_wf(name="generate_hcp_surfaces_wf"):
     inflate_surface = pe.Node(
         SurfaceGenerateInflated(iterations_scale_value=0.75),
         mem_gb=2,
-        omp_nthreads=omp_nthreads,
         name="inflate_surface",
     )
     workflow.connect([
@@ -614,7 +610,7 @@ def init_generate_hcp_surfaces_wf(name="generate_hcp_surfaces_wf"):
 
 
 @fill_doc
-def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
+def init_ants_xfm_to_fsl_wf(mem_gb, name="ants_xfm_to_fsl_wf"):
     """Modify ANTS-style fMRIPrep transforms to work with Connectome Workbench/FSL FNIRT.
 
     XXX: Does this only work if the template is MNI152NLin6Asym?
@@ -628,14 +624,12 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
 
             wf = init_ants_xfm_to_fsl_wf(
                 mem_gb=0.1,
-                omp_nthreads=1,
                 name="ants_xfm_to_fsl_wf",
             )
 
     Parameters
     ----------
     %(mem_gb)s
-    %(omp_nthreads)s
     %(name)s
         Default is "ants_xfm_to_fsl_wf".
 
@@ -679,7 +673,6 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         ),
         name="disassemble_h5",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     workflow.connect([(inputnode, disassemble_h5, [("anat_to_template_xfm", "in_file")])])
 
@@ -693,7 +686,6 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         ),
         name="disassemble_h5_inv",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     workflow.connect([(inputnode, disassemble_h5_inv, [("template_to_anat_xfm", "in_file")])])
 
@@ -726,7 +718,6 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         ),
         name="get_xyz_components",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     get_inv_xyz_components = pe.Node(
         C3d(
@@ -736,7 +727,6 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         ),
         name="get_inv_xyz_components",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     workflow.connect([
         (disassemble_h5, get_xyz_components, [("displacement_field", "in_file")]),
@@ -748,41 +738,35 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         niu.Select(index=[0]),
         name="select_x_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     select_inv_x_component = pe.Node(
         niu.Select(index=[0]),
         name="select_inv_x_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
 
     # Select y-component
     select_y_component = pe.Node(
         niu.Select(index=[1]),
         name="select_y_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     select_inv_y_component = pe.Node(
         niu.Select(index=[1]),
         name="select_inv_y_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
 
     # Select z-component
     select_z_component = pe.Node(
         niu.Select(index=[2]),
         name="select_z_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     select_inv_z_component = pe.Node(
         niu.Select(index=[2]),
         name="select_inv_z_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     workflow.connect([
         (get_xyz_components, select_x_component, [("out_files", "inlist")]),
@@ -800,13 +784,11 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         BinaryMath(expression="img * -1"),
         name="reverse_y_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     reverse_inv_y_component = pe.Node(
         BinaryMath(expression="img * -1"),
         name="reverse_inv_y_component",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     workflow.connect([
         (select_y_component, reverse_y_component, [("out", "in_file")]),
@@ -818,13 +800,11 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         niu.Merge(3),
         name="collect_new_components",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     collect_new_inv_components = pe.Node(
         niu.Merge(3),
         name="collect_new_inv_components",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     workflow.connect([
         (select_x_component, collect_new_components, [("out", "in1")]),
@@ -840,13 +820,11 @@ def init_ants_xfm_to_fsl_wf(mem_gb, omp_nthreads, name="ants_xfm_to_fsl_wf"):
         Merge(),
         name="remerge_warpfield",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     remerge_inv_warpfield = pe.Node(
         Merge(),
         name="remerge_inv_warpfield",
         mem_gb=mem_gb,
-        n_procs=omp_nthreads,
     )
     workflow.connect([
         (collect_new_components, remerge_warpfield, [("out", "in_files")]),

xcp_d/workflows/anatomical/volume.py

@@ -78,7 +78,6 @@ def init_postprocess_anat_wf(
     workflow = Workflow(name=name)
     output_dir = config.execution.xcp_d_dir
     input_type = config.workflow.input_type
-    omp_nthreads = config.nipype.omp_nthreads
 
     inputnode = pe.Node(
         niu.IdentityInterface(
@@ -168,7 +167,6 @@ def init_postprocess_anat_wf(
                 ),
                 name="warp_t1w_to_template",
                 mem_gb=2,
-                n_procs=omp_nthreads,
             )
             workflow.connect([
                 (inputnode, warp_t1w_to_template, [
@@ -189,7 +187,6 @@ def init_postprocess_anat_wf(
                 ),
                 name="warp_t2w_to_template",
                 mem_gb=2,
-                n_procs=omp_nthreads,
             )
             workflow.connect([
                 (inputnode, warp_t2w_to_template, [

xcp_d/workflows/bold/cifti.py

@@ -138,7 +138,6 @@ def init_postprocess_cifti_wf(
     dummy_scans = config.workflow.dummy_scans
     despike = config.workflow.despike
     atlases = config.execution.atlases
-    omp_nthreads = config.nipype.omp_nthreads
 
     TR = run_data["bold_metadata"]["RepetitionTime"]
 
@@ -217,7 +216,6 @@ def init_postprocess_cifti_wf(
         ConvertTo32(),
         name="downcast_data",
         mem_gb=mem_gbx["timeseries"],
-        n_procs=omp_nthreads,
     )
 
     workflow.connect([

xcp_d/workflows/bold/metrics.py

@@ -145,7 +145,6 @@ def init_alff_wf(
         ComputeALFF(TR=TR, low_pass=low_pass, high_pass=high_pass),
         mem_gb=mem_gb["resampled"],
         name="alff_compt",
-        n_procs=omp_nthreads,
     )
     workflow.connect([
         (inputnode, alff_compt, [
@@ -201,7 +200,6 @@ def init_alff_wf(
             smooth_data = pe.Node(
                 Smooth(fwhm=smoothing),
                 name="niftismoothing",
-                n_procs=omp_nthreads,
             )
             workflow.connect([
                 (alff_compt, smooth_data, [("alff", "in_file")]),
@@ -241,7 +239,6 @@ def init_alff_wf(
                 FixCiftiIntent(),
                 name="fix_cifti_intent",
                 mem_gb=mem_gb["resampled"],
-                n_procs=omp_nthreads,
             )
             workflow.connect([
                 (alff_compt, smooth_data, [("alff", "in_file")]),
@@ -330,39 +327,33 @@ def init_reho_cifti_wf(
         CiftiSeparateMetric(metric="CORTEX_LEFT", direction="COLUMN"),
         name="separate_lh",
         mem_gb=mem_gb["resampled"],
-        n_procs=omp_nthreads,
     )
     rh_surf = pe.Node(
         CiftiSeparateMetric(metric="CORTEX_RIGHT", direction="COLUMN"),
         name="separate_rh",
         mem_gb=mem_gb["resampled"],
-        n_procs=omp_nthreads,
     )
     subcortical_nifti = pe.Node(
         CiftiSeparateVolumeAll(direction="COLUMN"),
         name="separate_subcortical",
         mem_gb=mem_gb["resampled"],
-        n_procs=omp_nthreads,
     )
 
     # Calculate the reho by hemipshere
     lh_reho = pe.Node(
         SurfaceReHo(surf_hemi="L"),
         name="reho_lh",
         mem_gb=mem_gb["resampled"],
-        n_procs=omp_nthreads,
     )
     rh_reho = pe.Node(
         SurfaceReHo(surf_hemi="R"),
         name="reho_rh",
         mem_gb=mem_gb["resampled"],
-        n_procs=omp_nthreads,
     )
     subcortical_reho = pe.Node(
         ReHoNamePatch(neighborhood="vertices"),
         name="reho_subcortical",
         mem_gb=mem_gb["resampled"],
-        n_procs=omp_nthreads,
     )
 
     # Merge the surfaces and subcortical structures back into a CIFTI

xcp_d/workflows/bold/nifti.py

@@ -148,7 +148,6 @@ def init_postprocess_nifti_wf(
     dummy_scans = config.workflow.dummy_scans
     despike = config.workflow.despike
     atlases = config.execution.atlases
-    omp_nthreads = config.nipype.omp_nthreads
 
     TR = run_data["bold_metadata"]["RepetitionTime"]
 
@@ -227,7 +226,6 @@ def init_postprocess_nifti_wf(
         ConvertTo32(),
         name="downcast_data",
         mem_gb=mem_gbx["timeseries"],
-        n_procs=omp_nthreads,
     )
 
     workflow.connect([