Add training data

materialsvirtuallab · Aug 2, 2018 · 5131436 · 5131436
1 parent a5222f5
commit 5131436
Show file tree

Hide file tree

Showing 18 changed files with 89 additions and 0 deletions.
diff --git a/Cu/training/AIMD.json b/Cu/training/AIMD.json
diff --git a/Cu/training/Elastic.json b/Cu/training/Elastic.json
diff --git a/Cu/training/README.md b/Cu/training/README.md
@@ -0,0 +1,25 @@
+# Using the training data
+
+The training data is provided in json files, which can be read by any standard json parser. Each json file contains contains one data group (e.g., AIMD, elastic, surfaces, etc.) used in the training of the SNAP model.
+
+The data is provided as a list of dicts. Each dict contains one structure with data. The format of the json file is as follows:
+
+```json
+[
+    {   
+        "_id": "a string",
+        "params": {"Dict of parameters used for DFT calculations, e.g. pseudopotential, energy cutoff, etc."},
+        "group": "Cu",
+        "tags": "Md",
+        "structure": {"Dict representation of Structure as output by pymatgen"},
+        "num_atoms": 108,
+        "outputs": {
+            "energy": -619.08666252,
+            "forces": [],
+            "stress": []
+        }
+    },
+    ...
+]
+```
+
diff --git a/Cu/training/Surface.json b/Cu/training/Surface.json
diff --git a/Cu/training/Vacancy.json b/Cu/training/Vacancy.json
diff --git a/Ni-Mo/training/Mo_dopedwith_Ni.json b/Ni-Mo/training/Mo_dopedwith_Ni.json
diff --git a/Ni-Mo/training/Ni3Mo_AIMD.json b/Ni-Mo/training/Ni3Mo_AIMD.json
diff --git a/Ni-Mo/training/Ni3Mo_Elastic.json b/Ni-Mo/training/Ni3Mo_Elastic.json
diff --git a/Ni-Mo/training/Ni4Mo_AIMD.json b/Ni-Mo/training/Ni4Mo_AIMD.json
diff --git a/Ni-Mo/training/Ni4Mo_Elastic.json b/Ni-Mo/training/Ni4Mo_Elastic.json
diff --git a/Ni-Mo/training/Ni_dopedwith_Mo.json b/Ni-Mo/training/Ni_dopedwith_Mo.json
diff --git a/Ni-Mo/training/README.md b/Ni-Mo/training/README.md
@@ -0,0 +1,25 @@
+# Using the training data
+
+The training data is provided in json files, which can be read by any standard json parser. Each json file contains contains one data group (e.g., Ni3Mo_AIMD, Ni4Mo_elastic, Ni_dopedwith_Mo, etc.) used in the training of the SNAP model.
+
+The data is provided as a list of dicts. Each dict contains one structure with data. The format of the json file is as follows:
+
+```json
+[
+    {   
+        "_id": "a string",
+        "params": {"Dict of parameters used for DFT calculations, e.g. pseudopotential, energy cutoff, etc."},
+        "group": "Ni3Mo",
+        "tags": "Md",
+        "structure": {"Dict representation of Structure as output by pymatgen"},
+        "num_atoms": 108,
+        "outputs": {
+            "energy": -619.08666252,
+            "forces": [],
+            "stress": []
+        }
+    },
+    ...
+]
+```
+
diff --git a/Ni/training/.AIMD.json.swp b/Ni/training/.AIMD.json.swp
diff --git a/Ni/training/AIMD.json b/Ni/training/AIMD.json
diff --git a/Ni/training/Elastic.json b/Ni/training/Elastic.json
diff --git a/Ni/training/README.md b/Ni/training/README.md
@@ -0,0 +1,25 @@
+# Using the training data
+
+The training data is provided in json files, which can be read by any standard json parser. Each json file contains contains one data group (e.g., AIMD, elastic, surfaces, etc.) used in the training of the SNAP model.
+
+The data is provided as a list of dicts. Each dict contains one structure with data. The format of the json file is as follows:
+
+```json
+[
+    {   
+        "_id": "a string",
+        "params": {"Dict of parameters used for DFT calculations, e.g. pseudopotential, energy cutoff, etc."},
+        "group": "Ni",
+        "tags": "Md",
+        "structure": {"Dict representation of Structure as output by pymatgen"},
+        "num_atoms": 108,
+        "outputs": {
+            "energy": -619.08666252,
+            "forces": [],
+            "stress": []
+        }
+    },
+    ...
+]
+```
+
diff --git a/Ni/training/Surface.json b/Ni/training/Surface.json
diff --git a/Ni/training/Vacancy.json b/Ni/training/Vacancy.json