Bug Fixes (#100)

* develop branch * ToDict and FromDict methods added to classes * Interface routine fixed, added spacegroup tests, more queue jobs * Pycodestyle fix * Tests fix. * Update README.rst * Update README.rst * Create DatasetSummary.rst * Update README.rst * Update README.rst * Pydocstyle fix
usnistgov · Jul 23, 2020 · 5acac87 · 5acac87
1 parent a9322fc
commit 5acac87
Show file tree

Hide file tree

Showing 11 changed files with 257 additions and 289 deletions.
diff --git a/DatasetSummary.rst b/DatasetSummary.rst
@@ -0,0 +1,15 @@
+* A summary of the projects
+
+  ===============  =======================================================================
+  Projects          Brief description
+  ===============  =======================================================================
+  ``JARVIS-DFT``      Density functional theory calculation database for ~40000 3D and ~1000 2D materials. Some of the material-properties include: Heat of formation, Crystal-structural data using OptB88vdW, PBE, LDA functionals, Bandgaps using semi-local, meta-GGA, HSE06 and other beyond DFT methods, Electron and phonon-bandstructures, Elastic, Piezoelectric, Thermoelectric, Dielectric tensors, Exfoliation energies for low-diemnsional materials, Frequency dependent dielectric function, Absorption coefficients, Work-function for 2D materials, Infrared and Raman intensities, Electric field gradient, Magnetic moment, Solar-cell efficiencies, Scanning Tunneling Microscopy (STM) images, Topological spin-orbit spillage, converged k-point and plane wave cut-offs, Wannier-tight binding Hamiltonian parameters and more. The website for JARVIS-DFT: https://www.ctcms.nist.gov/~knc6/JVASP.html
+  ``JARVIS-FF``       Classical molecular dynamics calculation database for ~2000 3D materials with interatomic potential/force-fields. Some of the properties included in JARVIS-FF are energetics, elastic constants, surface energies, defect formations energies and phonon frequencies of materials. The website for JARVIS-FF: https://www.ctcms.nist.gov/~knc6/periodic.html
+  ``JARVIS-ML``       Machine learning prediction tools trained on the JARVIS-DFT data. Some of the ML-prediction models are for  Heat of formation, GGA/METAGGA bandgaps, Refractive indices, Bulk and shear modulus, Magnetic moment, Thermoelectric, Piezoelectric and Dielectric properties properties, Exfoliation energies, Solar-cell efficiency, and STM image classification. The website for JARVIS-ML: https://www.ctcms.nist.gov/jarvisml
+  ``JARVIS-Het.``     Heterostructure design tools for 2D materials in the JARVIS-DFT database. Some of the properties available are: work function, Band-alignment, and Heterostructure classification. JARVIS-Heterostructure website: https://www.ctcms.nist.gov/jarvish
+  ``JARVIS-PV``       Solar-cell/Photovoltaic cell design tools. Dataset is made available and the website will be available soon.
+  ``JARVIS-STM``      Scanning-tunneling microscopy images for 2D materials. Dataset is made available and the website will be available soon.
+  ``JARVIS-WTB``      Wannier Tight Binding Hamiltonian parameter dataset. Website: https://www.ctcms.nist.gov/jarviswtb 
+  ``JARVIS-EFG``      Electric field gradient dataset. Dataset will be made available and the website will be available soon.
+  ``Downloads``       Download raw metadat at: https://www.ctcms.nist.gov/~knc6/downloads.html
+  ===============  =======================================================================
diff --git a/README.rst b/README.rst
@@ -50,22 +50,16 @@ For more details, checkout our latest article:  `JARVIS: An Integrated Infrastru
    :target: https://jarvis.nist.gov/
 
 
+Some important features
+=======================================================================
 
-* A summary of the projects
-
-  ===============  =======================================================================
-  Projects          Brief description
-  ===============  =======================================================================
-  ``JARVIS-DFT``      Density functional theory calculation database for ~40000 3D and ~1000 2D materials. Some of the material-properties include: Heat of formation, Crystal-structural data using OptB88vdW, PBE, LDA functionals, Bandgaps using semi-local, meta-GGA, HSE06 and other beyond DFT methods, Electron and phonon-bandstructures, Elastic, Piezoelectric, Thermoelectric, Dielectric tensors, Exfoliation energies for low-diemnsional materials, Frequency dependent dielectric function, Absorption coefficients, Work-function for 2D materials, Infrared and Raman intensities, Electric field gradient, Magnetic moment, Solar-cell efficiencies, Scanning Tunneling Microscopy (STM) images, Topological spin-orbit spillage, converged k-point and plane wave cut-offs, Wannier-tight binding Hamiltonian parameters and more. The website for JARVIS-DFT: https://www.ctcms.nist.gov/~knc6/JVASP.html
-  ``JARVIS-FF``       Classical molecular dynamics calculation database for ~2000 3D materials with interatomic potential/force-fields. Some of the properties included in JARVIS-FF are energetics, elastic constants, surface energies, defect formations energies and phonon frequencies of materials. The website for JARVIS-FF: https://www.ctcms.nist.gov/~knc6/periodic.html
-  ``JARVIS-ML``       Machine learning prediction tools trained on the JARVIS-DFT data. Some of the ML-prediction models are for  Heat of formation, GGA/METAGGA bandgaps, Refractive indices, Bulk and shear modulus, Magnetic moment, Thermoelectric, Piezoelectric and Dielectric properties properties, Exfoliation energies, Solar-cell efficiency, and STM image classification. The website for JARVIS-ML: https://www.ctcms.nist.gov/jarvisml
-  ``JARVIS-Het.``     Heterostructure design tools for 2D materials in the JARVIS-DFT database. Some of the properties available are: work function, Band-alignment, and Heterostructure classification. JARVIS-Heterostructure website: https://www.ctcms.nist.gov/jarvish
-  ``JARVIS-PV``       Solar-cell/Photovoltaic cell design tools. Dataset is made available and the website will be available soon.
-  ``JARVIS-STM``      Scanning-tunneling microscopy images for 2D materials. Dataset is made available and the website will be available soon.
-  ``JARVIS-WTB``      Wannier Tight Binding Hamiltonian parameter dataset. Website: https://www.ctcms.nist.gov/jarviswtb 
-  ``JARVIS-EFG``      Electric field gradient dataset. Dataset will be made available and the website will be available soon.
-  ``Downloads``       Download raw metadat at: https://www.ctcms.nist.gov/~knc6/downloads.html
-  ===============  =======================================================================
+- **Software workflow tasks**:  VASP, Quantum Espresso, BoltzTrap, Wannier90, LAMMPS, Scikit-learn, TensorFlow, LightGBM.
+
+- **HPC clusters**: PBS and SLURM.
+
+- **Examples**: Notebooks and test scripts to explain the package.
+
+- **Available datasets**: `Summary of several datasets <https://github.com/usnistgov/jarvis/blob/master/DatasetSummary.rst>`__ .
 
 
 Installation

diff --git a/jarvis/analysis/interface/zur.py b/jarvis/analysis/interface/zur.py
@@ -2,7 +2,7 @@
 
 from itertools import product
 import numpy as np
-from jarvis.core.atoms import Atoms, add_atoms, fix_pbc, VacuumPadding
+from jarvis.core.atoms import add_atoms, fix_pbc
 from jarvis.core.lattice import Lattice
 
 
@@ -290,12 +290,12 @@ def get_factors(n):
 def make_interface(
     film="",
     subs="",
-    atol=0.01,
-    ltol=0.03,
+    atol=1,
+    ltol=0.05,
     max_area=500,
-    max_area_ratio_tol=0.09,
-    seperation=2.0,
-    vacuum=18.0,
+    max_area_ratio_tol=1.00,
+    seperation=3.0,
+    vacuum=8.0,
 ):
     """
     Use as main function for making interfaces/heterostructures.
@@ -308,19 +308,18 @@ def make_interface(
        subs: substrate/bottom/fixed material.
 
        seperation: minimum seperation between two.
+
+       vacuum: vacuum will be added on both sides.
+       So 2*vacuum will be added.
     """
     z = ZSLGenerator(
         max_area_ratio_tol=max_area_ratio_tol,
         max_area=max_area,
         max_length_tol=ltol,
         max_angle_tol=atol,
     )
-    film = fix_pbc(
-        film.center_around_origin([0, 0, 0])
-    )  # .get_string(cart=False)
-    subs = fix_pbc(
-        subs.center_around_origin([0, 0, 0])
-    )  # .get_string(cart=False)
+    film = fix_pbc(film.center_around_origin([0, 0, 0]))
+    subs = fix_pbc(subs.center_around_origin([0, 0, 0]))
     matches = list(z(film.lattice_mat[:2], subs.lattice_mat[:2], lowest=True))
     info = {}
     info["mismatch_u"] = "na"
@@ -387,112 +386,29 @@ def make_interface(
     film_bottom_z = min(np.array(film_scell.cart_coords)[:, 2])
     thickness_sub = abs(substrate_top_z - substrate_bot_z)
     thickness_film = abs(film_top_z - film_bottom_z)
-
-    min_distance = seperation + (thickness_sub)  # +thickness_film/ 2.0
     sub_z = (
-        (vacuum + thickness_sub + thickness_film)
+        (vacuum + substrate_top_z)
         * np.array(subs_scell.lattice_mat[2, :])
         / np.linalg.norm(subs_scell.lattice_mat[2, :])
-    )  # subs.lattice_mat[2, :]
+    )
+    shift_normal = (
+        sub_z / np.linalg.norm(sub_z) * seperation / np.linalg.norm(sub_z)
+    )
+    tmp = (
+        thickness_film / 2 + seperation + thickness_sub / 2
+    ) / np.linalg.norm(subs_scell.lattice_mat[2, :])
     shift_normal = (
-        sub_z / np.linalg.norm(sub_z) * min_distance / np.linalg.norm(sub_z)
+        tmp
+        * np.array(subs_scell.lattice_mat[2, :])
+        / np.linalg.norm(subs_scell.lattice_mat[2, :])
     )
     interface = add_atoms(
         film_scell, subs_scell, shift_normal
     ).center_around_origin([0, 0, 0.5])
-    info["interface"] = VacuumPadding(
-        atoms=interface, vacuum=vacuum
-    ).get_effective_2d_slab()
-    print("mismatch_u,mismatch_v", mismatch_u, mismatch_v)
-    if mismatch_u < 0 or mismatch_v < 0:
-        print("Maybe unphysical structure")
-    info["msg"] = "Maybe unphysical structure"
-    return info
-
-
-def mismatch_strts(film=[], subs=[], ltol=0.05, atol=10):
-    """
-    Return mismatch and other information as info dict.
-
-    Deprecated this module will be removes in the next version.
-    """
-    z = ZSLGenerator()
-    matches = list(z(film.lattice_mat[:2], subs.lattice_mat[:2], lowest=True))
-    info = {}
-    info["mismatch_u"] = "na"
-    info["mismatch_v"] = "na"
-    info["mismatch_angle"] = "na"
-    info["area1"] = "na"
-    info["area2"] = "na"
-    info["film_sl"] = film
-    info["subs_sl"] = subs
-
-    uv1 = matches[0]["sub_sl_vecs"]
-    uv2 = matches[0]["film_sl_vecs"]
-    u = np.array(uv1)
-    v = np.array(uv2)
-    a1 = u[0]
-    a2 = u[1]
-    b1 = v[0]
-    b2 = v[1]
-    mismatch_u = np.linalg.norm(b1) / np.linalg.norm(a1) - 1
-    mismatch_v = np.linalg.norm(b2) / np.linalg.norm(a2) - 1
-    angle1 = (
-        np.arccos(np.dot(a1, a2) / np.linalg.norm(a1) / np.linalg.norm(a2))
-        * 180
-        / np.pi
+    combined = interface.center(vacuum=vacuum).center_around_origin(
+        [0, 0, 0.5]
     )
-    angle2 = (
-        np.arccos(np.dot(b1, b2) / np.linalg.norm(b1) / np.linalg.norm(b2))
-        * 180
-        / np.pi
-    )
-    mismatch_angle = abs(angle1 - angle2)
-    area1 = np.linalg.norm(np.cross(a1, a2))
-    area2 = np.linalg.norm(np.cross(b1, b2))
-    uv_substrate = uv1
-    uv_mat2d = uv2
-    substrate_latt = Lattice(
-        np.array(
-            [uv_substrate[0][:], uv_substrate[1][:], subs.lattice_mat[2, :]]
-        )
-    )
-    film_norm = np.linalg.norm(film.lattice_mat[2, :])
-    mat2d_fake_c = film.lattice_mat[2, :] / film_norm * 5.0
-    mat2d_latt = Lattice(
-        np.array([uv_mat2d[0][:], uv_mat2d[1][:], mat2d_fake_c])
-    )
-    mat2d_latt_fake = Lattice(
-        np.array(
-            [film.lattice_mat[0, :], film.lattice_mat[1, :], mat2d_fake_c]
-        )
-    )
-    _, __, scell = subs.lattice.find_matches(
-        substrate_latt, ltol=ltol, atol=atol
-    )
-    scell[2] = np.array([0, 0, 1])
-    subs = subs.make_supercell_matrix(scell)
-    _, __, scell = mat2d_latt_fake.find_matches(mat2d_latt, ltol=0.05, atol=1)
-    scell[2] = np.array([0, 0, 1])
-    film = film.make_supercell_matrix(scell)
-
-    lmap = Lattice(
-        np.array(
-            [
-                subs.lattice_mat[0, :],
-                subs.lattice_mat[1, :],
-                film.lattice_mat[2, :],
-            ]
-        )
-    )
-    film.lattice = lmap
-    info["mismatch_u"] = mismatch_u
-    info["mismatch_v"] = mismatch_v
-    info["mismatch_angle"] = mismatch_angle
-    info["area1"] = area1
-    info["area2"] = area2
-    info["film_sl"] = film
-    info["subs_sl"] = subs
+    info["interface"] = combined
     return info
 
 
@@ -524,42 +440,6 @@ def get_hetero_type(A={}, B={}):
     return int_type, stack
 
 
-def get_hetero(film, substrate, seperation=3.5):
-    """Generate heterostructure, deprecated."""
-    substrate_top_z = max(np.array(substrate.cart_coords)[:, 2])
-    # substrate_bot_z = min(np.array(substrate.cart_coords)[:, 2])
-    # film_bottom = min(np.array(film.cart_coords)[:, 2])
-    # film_top = max(np.array(film.cart_coords)[:, 2])
-    sub_z = substrate.lattice_mat[2, :]
-    origin = np.array([0, 0, substrate_top_z])
-    shift_normal = sub_z / np.linalg.norm(sub_z) * seperation
-    # thickness_sub = abs(substrate_top_z - substrate_bot_z)
-    # thickness_film = abs(film_top - film_bottom)
-    new_coords = []
-    lattice_mat = substrate.lattice_mat
-    elements = []
-    for i in substrate.cart_coords:
-        new_coords.append(i)
-    for i in substrate.elements:
-        elements.append(i)
-
-    for i in film.elements:
-        elements.append(i)
-    for i in film.cart_coords:
-        tmp = i
-        tmp = tmp + origin + shift_normal
-        new_coords.append(tmp)
-
-    interface = Atoms(
-        lattice_mat=lattice_mat,
-        elements=elements,
-        coords=new_coords,
-        cartesian=True,
-    )
-
-    return interface
-
-
 """
 if __name__ == "__main__":
     s1 = Poscar.from_file(

diff --git a/jarvis/core/atoms.py b/jarvis/core/atoms.py
@@ -115,7 +115,7 @@ def to_dict(self):
         d = OrderedDict()
         d["lattice_mat"] = self.lattice_mat.tolist()
         d["coords"] = np.array(self.coords).tolist()
-        d["elements"] = self.elements
+        d["elements"] = np.array(self.elements).tolist()
         d["abc"] = self.lattice.lat_lengths()
         d["angles"] = self.lattice.lat_angles()
         d["cartesian"] = self.cartesian

diff --git a/jarvis/io/boltztrap/outputs.py b/jarvis/io/boltztrap/outputs.py
@@ -30,15 +30,14 @@ def __init__(
             condtens_fixdoping = self.read_condtens_fixdoping()
         self.condtens_fixdoping = condtens_fixdoping
         if halltens_fixdoping == []:
-            halltens_fixdoping = self.read_condtens_fixdoping()
+            halltens_fixdoping = self.read_halltens_fixdoping()
         self.halltens_fixdoping = halltens_fixdoping
 
     def to_dict(self):
         """Return output as a dictionary."""
         d = OrderedDict()
         d["path"] = self.path
         d["outtrans_data"] = self.outtrans_data
-
         d["intrans_data"] = self.intrans_data
         d["halltens_fixdoping"] = self.halltens_fixdoping
         d["condtens_fixdoping"] = self.condtens_fixdoping

diff --git a/jarvis/io/vasp/inputs.py b/jarvis/io/vasp/inputs.py
@@ -304,7 +304,7 @@ def __init__(
         elements=[],
         pot_type="POT_GGA_PAW_PBE",
         pot_json_path="",
-        potcar_strings={},
+        potcar_strings=[],
     ):
         """
         Initialize the Potcar class.
@@ -329,21 +329,24 @@ def __init__(
         self._potcar_strings = potcar_strings
         self._pot_json_path = pot_json_path
 
-        if self._potcar_strings == {}:
+        if self._potcar_strings == []:
             pot_json_file = str(
                 os.path.join(os.path.dirname(__file__), "default_potcars.json")
             )
             self._pot_json_path = pot_json_file
             pot_json = open(pot_json_file, "r")
             pot_json_selected = json.load(pot_json)
             pot_json.close()
-            potcar_strings = OrderedDict()
+            potcar_strings = []  # OrderedDict()
             for i in self._elements:
-                for j, k in pot_json_selected.items():
-                    if i == j:
-                        potcar_strings.setdefault(i, k)
+                potcar_strings.append(pot_json_selected[i])
+                # for j, k in pot_json_selected.items():
+                #    if i == j:
+                #        potcar_strings.setdefault(i, k)
             self._potcar_strings = potcar_strings
-            if len(self._elements) != len(self._potcar_strings.keys()):
+            print("self._elements", self._elements)
+            print("self._potcar_strings", self._potcar_strings)
+            if len(self._elements) != len(self._potcar_strings):
                 raise ValueError(
                     "Number of elements is not same as potcar_strings",
                     self._elements,
@@ -353,12 +356,8 @@ def __init__(
             pot_json = open(self._pot_json_path, "r")
             pot_json_selected = json.load(pot_json)
             pot_json.close()
-            potcar_strings = OrderedDict()
-            for i, j in pot_json_selected.items():
-                if i in self._elements:
-                    potcar_strings.setdefault(i, j)
             self._potcar_strings = potcar_strings
-            if len(self._elements) != len(self._potcar_strings.keys()):
+            if len(self._elements) != len(self._potcar_strings):
                 msg = "Number of elements not same as potcar_strings"
                 raise ValueError(msg)
 
@@ -375,10 +374,10 @@ def from_dict(self, d={}):
     def to_dict(self):
         """Convert to a dictionary."""
         d = OrderedDict()
-        d["elements"] = self._elements
+        d["elements"] = np.array(self._elements).tolist()
         d["pot_type"] = self._pot_type
         d["pot_json_path"] = self._pot_json_path
-        d["potcar_strings"] = self._potcar_strings
+        d["potcar_strings"] = np.array(self._potcar_strings).tolist()
         return d
 
     def catenate_potcar_files(
@@ -397,7 +396,7 @@ def list_potcar_files(self):
         vasp_dir = str(os.environ["JARVIS_VASP_PSP_DIR"])
         vasp_psp_dir = str(os.path.join(vasp_dir, self._pot_type))
         potcar_strings = self._potcar_strings
-        for i, j in potcar_strings.items():
+        for j in potcar_strings:
             tmp = os.path.join(vasp_psp_dir, j, "POTCAR")
             pot_files.append(tmp)
         return pot_files