Pseudopotential and band update (#113)

* Pseudopotential and band update Section 3.2 gets a version of the SSSP family that is not compatible with the bandstructure workflow in the current version. (updated there) Explained where to get the Upfdata, how to check if it is correct (Al case) and how to run the workflow with the 1.0 SSSP family * Addressing the correction to Bands in #113 * Errors and correction in text and notebook * fix further download typo
aiidateam · May 31, 2019 · d4b01bd · d4b01bd
1 parent ea53a57
commit d4b01bd
Show file tree

Hide file tree

Showing 3 changed files with 58 additions and 10 deletions.
diff --git a/docs/pages/2019_MARVEL_Psik_MaX/notebooks/bandstructure.ipynb b/docs/pages/2019_MARVEL_Psik_MaX/notebooks/bandstructure.ipynb
@@ -103,10 +103,25 @@
     "* Refine the symmetry of the relaxed structure to ensure the primitive cell is used and run a self-consistent field calculation on it\n",
     "* Run a non self-consistent field band structure calculation along a path of high symmetry k-points determined by [seekpath](http://materialscloud.org/tools/seekpath)\n",
     "\n",
-    "Numerical parameters for the default 'theos-ht-1.0' protocol are determined as follows:\n",
-    "* Suitable pseudopotentials and energy cutoffs are automatically searched from the [SSSP library](http://materialscloud.org/sssp) installed on your machine  (it uses the efficiency version 1.1)\n",
+    "Numerical parameters for the default 'theos-ht-1.0' protocol are determined as follows: \n",
+    "* Suitable pseudopotentials and energy cutoffs are automatically searched from the [SSSP library](http://materialscloud.org/sssp) installed on your machine  (it uses the efficiency version 1.1).\n",
     "* K-point mesh is selected to have a minimum k-point density of 0.2 Å<sup>-1</sup>\n",
-    "* A Marzari-Vanderbilt smearing of 0.02 Ry is used for the electronic occupations"
+    "* A Marzari-Vanderbilt smearing of 0.02 Ry is used for the electronic occupations\n",
+    "\n",
+    "In case the pseudopotentials are not installed, they can be downloaded in a terminal as:\n",
+    "\n",
+    "    wget https://archive.materialscloud.org/file/2018.0001/v3/SSSP_efficiency_pseudos.tar.gz\n",
+    "    tar -zxvf SSSP_efficiency_pseudos.tar.gz\n",
+    "    \n",
+    "The protocol looks for a UPF file with a specific hash code, that is unique for each different file. \n",
+    "You can check that you have the right\n",
+    "one by performing a search in the database:\n",
+    "\n",
+    "    qb=QueryBuilder()\n",
+    "    qb.append(UpfData, filters={'attributes.md5':{'==':'cfc449ca30b5f3223ec38ddd88ac046d'}})\n",
+    "    len(qb.all())\n",
+    "\n",
+    "'md5' is a searchable attribute of the pseudopotential data object."
    ]
   },
   {
@@ -138,6 +153,42 @@
     "    **results['seekpath_parameters'].get_dict()))"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "If you want to use a different pseudopotential family (or version of the family) (for instance [SSSP v1.0](https://archive.materialscloud.org/file/2018.0001/v1/SSSP_efficiency_pseudos.tar.gz) instead of the default SSSP v1.1) you can pass an additional parameter when calling the WorkChain, as follows:\n",
+    "   \n",
+    "    protocol = Dict(dict={\n",
+    "     'name':'theos-ht-1.0', \n",
+    "     'modifiers': {\n",
+    "       'pseudo' : 'SSSP-efficiency-1.0'\n",
+    "       }\n",
+    "     })\n",
+    "\n",
+    "(note that only some values are accepted for pseudo, that you can find [here](https://github.com/aiidateam/aiida-quantumespresso/blob/b02250146576eb573ccb45d05047075f54853f9d/aiida_quantumespresso/utils/protocols/pw.py#L24))."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# This will take approximately 6 minutes on the tutorial AWS (for Al)\n",
+    "results = launch.run(\n",
+    "    PwBandStructureWorkChain,\n",
+    "    code=code,\n",
+    "    structure=structure_Al,\n",
+    "    protocol=Dict(dict={\n",
+    "        'name':'theos-ht-1.0',\n",
+    "        'modifiers': {\n",
+    "            'pseudo':'SSSP-efficiency-1.0'\n",
+    "        }\n",
+    "    })\n",
+    ")"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,

diff --git a/docs/pages/2019_MARVEL_Psik_MaX/sections/bands.rst b/docs/pages/2019_MARVEL_Psik_MaX/sections/bands.rst
@@ -14,12 +14,9 @@ pseudopotentials, energy cutoffs, k-point meshes, high-symmetry k-point paths,
 and performing the various calculation steps -- are performed automatically by
 the ``WorkChain``.
 
-The demonstration of the ``WorkChain`` will be performed in a Jupyter
-notebook. To run it, follow the instructions that were given for the
-``QueryBuilder`` notebook in :numref:`querybuilder`. The only difference
-is that instead of selecting the notebook in the ``querybuilder``
-directory, go to ``pw/bandstructure`` and choose the ``bandstructure.ipynb``
-notebook. There you will find some example structures that are loaded from
+The demonstration of the WorkChain will be performed in a Jupyter notebook, 
+that you can :download:`download from here <../notebooks/bandstructure.ipynb>`.
+There you will find some example structures that are loaded from the
 Crystallography Open Database (COD), using the COD-importer integrated in
 AiiDA.
 

diff --git a/docs/pages/2019_MARVEL_Psik_MaX/sections/verdi_shell.rst b/docs/pages/2019_MARVEL_Psik_MaX/sections/verdi_shell.rst
@@ -398,7 +398,7 @@ SSSP pseudopotentials via the commands:
 
 .. code:: bash
 
-    wget https://archive.materialscloud.org/file/2018.0001/v1/SSSP_efficiency_pseudos.tar.gz
+    wget https://archive.materialscloud.org/file/2018.0001/v3/SSSP_efficiency_pseudos.tar.gz
     tar -zxvf SSSP_efficiency_pseudos.tar.gz
 
 Then you can upload the whole set of pseudopotentials to AiiDA by using the