class MPDataRetrieval:
"""
MPDataRetrieval is used to retrieve data from the Materials Project database, print the results, and convert them
into an indexed/unindexed Pandas dataframe.
"""
def __init__(self, api_key=None):
"""
Args:
api_key: (str) Your Citrine API key, or None if you've set the CITRINE_KEY environment variable
Returns: None
"""
self.mprester = MPRester(api_key=api_key)
def get_dataframe(self, criteria, properties, mp_decode=False, index_mpid=True):
"""
Gets data from MP in a dataframe format.
See API docs at https://materialsproject.org/wiki/index.php/The_Materials_API for more details.
Args:
criteria: (str/dict) Criteria of the query as a string or mongo-style dict. If string, it supports a
powerful but simple string criteria. E.g., "Fe2O3" means search for materials with reduced_formula
Fe2O3. Wild cards are also supported. E.g., "\*2O" means get all materials whose formula can be formed
as \*2O, e.g., Li2O, K2O, etc.
Other syntax examples:
mp-1234: Interpreted as a Materials ID.
Fe2O3 or \*2O3: Interpreted as reduced formulas.
Li-Fe-O or \*-Fe-O: Interpreted as chemical systems.
You can mix and match with spaces, which are interpreted as "OR". E.g. "mp-1234 FeO" means query for
all compounds with reduced formula FeO or with materials_id mp-1234.
Using a full dict syntax, even more powerful queries can be constructed. For example,
{"elements":{"$in":["Li", "Na", "K"], "$all": ["O"]}, "nelements":2} selects all Li, Na and K oxides.
{"band_gap": {"$gt": 1}} selects all materials with band gaps greater than 1 eV.
properties: (list) Properties to request for as a list. For example,
["formula", "formation_energy_per_atom"] returns the formula and formation energy per atom.
mp_decode: (bool) Whether to do a decoding to a Pymatgen object where possible. In some cases, it might be
useful to just get the raw python dict, i.e., set to False.
index_mpid: (bool) Whether to set the materials_id as the dataframe index.
Returns: A Pandas dataframe object
"""
if index_mpid and "material_id" not in properties:
properties.append("material_id")
data = self.mprester.query(criteria, properties, mp_decode)
df = pd.DataFrame(data, columns=properties)
if index_mpid:
df = df.set_index("material_id")
return df
class MPResterTest(unittest.TestCase):
def setUp(self):
self.rester = MPRester()
def test_get_data(self):
props = ["energy", "energy_per_atom", "formation_energy_per_atom",
"nsites", "unit_cell_formula", "pretty_formula", "is_hubbard",
"elements", "nelements", "e_above_hull", "hubbards",
"is_compatible", "task_ids",
"density", "icsd_ids", "total_magnetization"]
# unicode literals have been reintroduced in py>3.2
expected_vals = [-191.33812137, -6.833504334642858, -2.551358929370749,
28, {k: v for k, v in {'P': 4, 'Fe': 4, 'O': 16, 'Li': 4}.items()},
"LiFePO4", True, ['Li', 'O', 'P', 'Fe'], 4, 0.0,
{k: v for k, v in {'Fe': 5.3, 'Li': 0.0, 'O': 0.0, 'P': 0.0}.items()}, True,
[u'mp-601412', u'mp-19017', u'mp-796535', u'mp-797820',
u'mp-540081', u'mp-797269'],
3.4662026991351147,
[159107, 154117, 160776, 99860, 181272, 166815,
260571, 92198, 165000, 155580, 38209, 161479, 153699,
260569, 260570, 200155, 260572, 181341, 181342,
72545, 56291, 97764, 162282, 155635],
16.0002716]
for (i, prop) in enumerate(props):
if prop not in ['hubbards', 'unit_cell_formula', 'elements',
'icsd_ids', 'task_ids']:
val = self.rester.get_data("mp-19017", prop=prop)[0][prop]
self.assertAlmostEqual(expected_vals[i], val)
elif prop in ["elements", "icsd_ids", "task_ids"]:
self.assertEqual(set(expected_vals[i]),
set(self.rester.get_data("mp-19017",
prop=prop)[0][prop]))
else:
self.assertEqual(expected_vals[i],
self.rester.get_data("mp-19017",
prop=prop)[0][prop])
props = ['structure', 'initial_structure', 'final_structure', 'entry']
for prop in props:
print(prop)
obj = self.rester.get_data("mp-19017", prop=prop)[0][prop]
if prop.endswith("structure"):
self.assertIsInstance(obj, Structure)
elif prop == "entry":
obj = self.rester.get_data("mp-19017", prop=prop)[0][prop]
self.assertIsInstance(obj, ComputedEntry)
#Test chemsys search
data = self.rester.get_data('Fe-Li-O', prop='unit_cell_formula')
self.assertTrue(len(data) > 1)
elements = {Element("Li"), Element("Fe"), Element("O")}
for d in data:
self.assertTrue(
set(Composition(d['unit_cell_formula']).elements).issubset(
elements))
self.assertRaises(MPRestError, self.rester.get_data, "Fe2O3",
"badmethod")
def test_get_materials_id_from_task_id(self):
self.assertEqual(self.rester.get_materials_id_from_task_id(
"mp-540081"), "mp-19017")
def test_get_materials_id_references(self):
# nosetests pymatgen/matproj/tests/test_rest.py:MPResterTest.test_get_materials_id_references
m = MPRester()
data = m.get_materials_id_references('mp-123')
self.assertTrue(len(data) > 1000)
def test_find_structure(self):
# nosetests pymatgen/matproj/tests/test_rest.py:MPResterTest.test_find_structure
m = MPRester()
ciffile = os.path.join(test_dir, 'Fe3O4.cif')
data = m.find_structure(ciffile)
self.assertTrue(len(data) > 1)
s = CifParser(ciffile).get_structures()[0]
data = m.find_structure(s)
self.assertTrue(len(data) > 1)
def test_get_entries_in_chemsys(self):
syms = ["Li", "Fe", "O"]
entries = self.rester.get_entries_in_chemsys(syms)
elements = set([Element(sym) for sym in syms])
for e in entries:
self.assertIsInstance(e, ComputedEntry)
self.assertTrue(set(e.composition.elements).issubset(elements))
def test_get_structure_by_material_id(self):
s1 = self.rester.get_structure_by_material_id("mp-1")
self.assertEqual(s1.formula, "Cs1")
def test_get_entry_by_material_id(self):
e = self.rester.get_entry_by_material_id("mp-19017")
self.assertIsInstance(e, ComputedEntry)
self.assertTrue(e.composition.reduced_formula, "LiFePO4")
def test_query(self):
criteria = {'elements': {'$in': ['Li', 'Na', 'K'], '$all': ['O']}}
props = ['pretty_formula', 'energy']
data = self.rester.query(criteria=criteria, properties=props)
self.assertTrue(len(data) > 6)
data = self.rester.query(criteria="*2O", properties=props)
self.assertGreaterEqual(len(data), 52)
self.assertIn("Li2O", (d["pretty_formula"] for d in data))
def test_get_exp_thermo_data(self):
data = self.rester.get_exp_thermo_data("Fe2O3")
self.assertTrue(len(data) > 0)
for d in data:
self.assertEqual(d.formula, "Fe2O3")
def test_get_dos_by_id(self):
dos = self.rester.get_dos_by_material_id("mp-2254")
self.assertIsInstance(dos, CompleteDos)
def test_get_bandstructure_by_material_id(self):
bs = self.rester.get_bandstructure_by_material_id("mp-2254")
self.assertIsInstance(bs, BandStructureSymmLine)
def test_get_structures(self):
structs = self.rester.get_structures("Mn3O4")
self.assertTrue(len(structs) > 0)
def test_get_entries(self):
entries = self.rester.get_entries("TiO2")
self.assertTrue(len(entries) > 1)
for e in entries:
self.assertEqual(e.composition.reduced_formula, "TiO2")
entries = self.rester.get_entries("TiO2", inc_structure="final")
self.assertTrue(len(entries) > 1)
for e in entries:
self.assertEqual(e.structure.composition.reduced_formula, "TiO2")
all_entries = self.rester.get_entries("Fe", compatible_only=False)
entries = self.rester.get_entries("Fe", compatible_only=True)
self.assertTrue(len(entries) < len(all_entries))
entries = self.rester.get_entries("Fe", compatible_only=True,
property_data=["cif"])
self.assertIn("cif", entries[0].data)
def test_get_exp_entry(self):
entry = self.rester.get_exp_entry("Fe2O3")
self.assertEqual(entry.energy, -825.5)
def test_submit_query_delete_snl(self):
s = Structure([[5, 0, 0], [0, 5, 0], [0, 0, 5]], ["Fe"], [[0, 0, 0]])
# d = self.rester.submit_snl(
# [s, s], remarks=["unittest"],
# authors="Test User <*****@*****.**>")
# self.assertEqual(len(d), 2)
# data = self.rester.query_snl({"about.remarks": "unittest"})
# self.assertEqual(len(data), 2)
# snlids = [d["_id"] for d in data]
# self.rester.delete_snl(snlids)
# data = self.rester.query_snl({"about.remarks": "unittest"})
# self.assertEqual(len(data), 0)
def test_get_stability(self):
entries = self.rester.get_entries_in_chemsys(["Fe", "O"])
modified_entries = []
for entry in entries:
# Create modified entries with energies that are 0.01eV higher
# than the corresponding entries.
if entry.composition.reduced_formula == "Fe2O3":
modified_entries.append(
ComputedEntry(entry.composition,
entry.uncorrected_energy + 0.01,
parameters=entry.parameters,
entry_id="mod_{}".format(entry.entry_id)))
rest_ehulls = self.rester.get_stability(modified_entries)
all_entries = entries + modified_entries
compat = MaterialsProjectCompatibility()
all_entries = compat.process_entries(all_entries)
pd = PhaseDiagram(all_entries)
a = PDAnalyzer(pd)
for e in all_entries:
if str(e.entry_id).startswith("mod"):
for d in rest_ehulls:
if d["entry_id"] == e.entry_id:
data = d
break
self.assertAlmostEqual(a.get_e_above_hull(e),
data["e_above_hull"])
def test_get_reaction(self):
rxn = self.rester.get_reaction(["Li", "O"], ["Li2O"])
self.assertIn("Li2O", rxn["Experimental_references"])
def test_get_substrates(self):
substrate_data = self.rester.get_substrates('mp-123', 5, [1, 0, 0])
substrates = [sub_dict['sub_id'] for sub_dict in substrate_data]
self.assertIn("mp-2534", substrates)
def test_parse_criteria(self):
crit = MPRester.parse_criteria("mp-1234 Li-*")
self.assertIn("Li-O", crit["$or"][1]["chemsys"]["$in"])
self.assertIn({"task_id": "mp-1234"}, crit["$or"])
crit = MPRester.parse_criteria("Li2*")
self.assertIn("Li2O", crit["pretty_formula"]["$in"])
self.assertIn("Li2I", crit["pretty_formula"]["$in"])
self.assertIn("CsLi2", crit["pretty_formula"]["$in"])
crit = MPRester.parse_criteria("Li-*-*")
self.assertIn("Li-Re-Ru", crit["chemsys"]["$in"])
self.assertNotIn("Li-Li", crit["chemsys"]["$in"])
comps = MPRester.parse_criteria("**O3")["pretty_formula"]["$in"]
for c in comps:
self.assertEqual(len(Composition(c)), 3, "Failed in %s" % c)
chemsys = MPRester.parse_criteria("{Fe,Mn}-O")["chemsys"]["$in"]
self.assertEqual(len(chemsys), 2)
comps = MPRester.parse_criteria("{Fe,Mn,Co}O")["pretty_formula"]["$in"]
self.assertEqual(len(comps), 3, comps)
#Let's test some invalid symbols
self.assertRaises(ValueError, MPRester.parse_criteria, "li-fe")
self.assertRaises(ValueError, MPRester.parse_criteria, "LO2")
crit = MPRester.parse_criteria("POPO2")
self.assertIn("P2O3", crit["pretty_formula"]["$in"])
class MPResterTest(unittest.TestCase):
def setUp(self):
self.rester = MPRester()
def test_get_data(self):
props = [
"energy", "energy_per_atom", "formation_energy_per_atom", "nsites",
"unit_cell_formula", "pretty_formula", "is_hubbard", "elements",
"nelements", "e_above_hull", "hubbards", "is_compatible",
"task_ids", "density", "icsd_ids", "total_magnetization"
]
# unicode literals have been reintroduced in py>3.2
expected_vals = [
-191.33812137, -6.833504334642858, -2.551358929370749, 28,
{k: v
for k, v in {
'P': 4,
'Fe': 4,
'O': 16,
'Li': 4
}.items()}, "LiFePO4", True, ['Li', 'O', 'P', 'Fe'], 4, 0.0,
{
k: v
for k, v in {
'Fe': 5.3,
'Li': 0.0,
'O': 0.0,
'P': 0.0
}.items()
}, True,
[
u'mp-601412', u'mp-19017', u'mp-796535', u'mp-797820',
u'mp-540081', u'mp-797269'
], 3.4662026991351147,
[
159107, 154117, 160776, 99860, 181272, 166815, 260571, 92198,
165000, 155580, 38209, 161479, 153699, 260569, 260570, 200155,
260572, 181341, 181342, 72545, 56291, 97764, 162282, 155635
], 16.0002716
]
for (i, prop) in enumerate(props):
if prop not in [
'hubbards', 'unit_cell_formula', 'elements', 'icsd_ids',
'task_ids'
]:
val = self.rester.get_data("mp-19017", prop=prop)[0][prop]
self.assertAlmostEqual(expected_vals[i], val)
elif prop in ["elements", "icsd_ids", "task_ids"]:
self.assertEqual(
set(expected_vals[i]),
set(self.rester.get_data("mp-19017", prop=prop)[0][prop]))
else:
self.assertEqual(
expected_vals[i],
self.rester.get_data("mp-19017", prop=prop)[0][prop])
props = ['structure', 'initial_structure', 'final_structure', 'entry']
for prop in props:
obj = self.rester.get_data("mp-19017", prop=prop)[0][prop]
if prop.endswith("structure"):
self.assertIsInstance(obj, Structure)
elif prop == "entry":
obj = self.rester.get_data("mp-19017", prop=prop)[0][prop]
self.assertIsInstance(obj, ComputedEntry)
#Test chemsys search
data = self.rester.get_data('Fe-Li-O', prop='unit_cell_formula')
self.assertTrue(len(data) > 1)
elements = {Element("Li"), Element("Fe"), Element("O")}
for d in data:
self.assertTrue(
set(Composition(
d['unit_cell_formula']).elements).issubset(elements))
self.assertRaises(MPRestError, self.rester.get_data, "Fe2O3",
"badmethod")
def test_get_materials_id_from_task_id(self):
self.assertEqual(
self.rester.get_materials_id_from_task_id("mp-540081"), "mp-19017")
def test_get_materials_id_references(self):
# nosetests pymatgen/matproj/tests/test_rest.py:MPResterTest.test_get_materials_id_references
m = MPRester()
data = m.get_materials_id_references('mp-123')
self.assertTrue(len(data) > 1000)
def test_find_structure(self):
# nosetests pymatgen/matproj/tests/test_rest.py:MPResterTest.test_find_structure
m = MPRester()
ciffile = os.path.join(test_dir, 'Fe3O4.cif')
data = m.find_structure(ciffile)
self.assertTrue(len(data) > 1)
s = CifParser(ciffile).get_structures()[0]
data = m.find_structure(s)
self.assertTrue(len(data) > 1)
def test_get_entries_in_chemsys(self):
syms = ["Li", "Fe", "O"]
entries = self.rester.get_entries_in_chemsys(syms)
elements = set([Element(sym) for sym in syms])
for e in entries:
self.assertIsInstance(e, ComputedEntry)
self.assertTrue(set(e.composition.elements).issubset(elements))
def test_get_structure_by_material_id(self):
s1 = self.rester.get_structure_by_material_id("mp-1")
self.assertEqual(s1.formula, "Cs1")
def test_get_entry_by_material_id(self):
e = self.rester.get_entry_by_material_id("mp-19017")
self.assertIsInstance(e, ComputedEntry)
self.assertTrue(e.composition.reduced_formula, "LiFePO4")
def test_query(self):
criteria = {'elements': {'$in': ['Li', 'Na', 'K'], '$all': ['O']}}
props = ['pretty_formula', 'energy']
data = self.rester.query(criteria=criteria, properties=props)
self.assertTrue(len(data) > 6)
data = self.rester.query(criteria="*2O", properties=props)
self.assertGreaterEqual(len(data), 52)
self.assertIn("Li2O", (d["pretty_formula"] for d in data))
def test_get_exp_thermo_data(self):
data = self.rester.get_exp_thermo_data("Fe2O3")
self.assertTrue(len(data) > 0)
for d in data:
self.assertEqual(d.formula, "Fe2O3")
def test_get_dos_by_id(self):
dos = self.rester.get_dos_by_material_id("mp-2254")
self.assertIsInstance(dos, CompleteDos)
def test_get_bandstructure_by_material_id(self):
bs = self.rester.get_bandstructure_by_material_id("mp-2254")
self.assertIsInstance(bs, BandStructureSymmLine)
def test_get_structures(self):
structs = self.rester.get_structures("Mn3O4")
self.assertTrue(len(structs) > 0)
def test_get_entries(self):
entries = self.rester.get_entries("TiO2")
self.assertTrue(len(entries) > 1)
for e in entries:
self.assertEqual(e.composition.reduced_formula, "TiO2")
entries = self.rester.get_entries("TiO2", inc_structure="final")
self.assertTrue(len(entries) > 1)
for e in entries:
self.assertEqual(e.structure.composition.reduced_formula, "TiO2")
all_entries = self.rester.get_entries("Fe", compatible_only=False)
entries = self.rester.get_entries("Fe", compatible_only=True)
self.assertTrue(len(entries) < len(all_entries))
entries = self.rester.get_entries("Fe",
compatible_only=True,
property_data=["cif"])
self.assertIn("cif", entries[0].data)
def test_get_exp_entry(self):
entry = self.rester.get_exp_entry("Fe2O3")
self.assertEqual(entry.energy, -825.5)
def test_submit_query_delete_snl(self):
s = Structure([[5, 0, 0], [0, 5, 0], [0, 0, 5]], ["Fe"], [[0, 0, 0]])
# d = self.rester.submit_snl(
# [s, s], remarks=["unittest"],
# authors="Test User <*****@*****.**>")
# self.assertEqual(len(d), 2)
# data = self.rester.query_snl({"about.remarks": "unittest"})
# self.assertEqual(len(data), 2)
# snlids = [d["_id"] for d in data]
# self.rester.delete_snl(snlids)
# data = self.rester.query_snl({"about.remarks": "unittest"})
# self.assertEqual(len(data), 0)
def test_get_stability(self):
entries = self.rester.get_entries_in_chemsys(["Fe", "O"])
modified_entries = []
for entry in entries:
# Create modified entries with energies that are 0.01eV higher
# than the corresponding entries.
if entry.composition.reduced_formula == "Fe2O3":
modified_entries.append(
ComputedEntry(entry.composition,
entry.uncorrected_energy + 0.01,
parameters=entry.parameters,
entry_id="mod_{}".format(entry.entry_id)))
rest_ehulls = self.rester.get_stability(modified_entries)
all_entries = entries + modified_entries
compat = MaterialsProjectCompatibility()
all_entries = compat.process_entries(all_entries)
pd = PhaseDiagram(all_entries)
a = PDAnalyzer(pd)
for e in all_entries:
if str(e.entry_id).startswith("mod"):
for d in rest_ehulls:
if d["entry_id"] == e.entry_id:
data = d
break
self.assertAlmostEqual(a.get_e_above_hull(e),
data["e_above_hull"])
def test_get_reaction(self):
rxn = self.rester.get_reaction(["Li", "O"], ["Li2O"])
self.assertIn("Li2O", rxn["Experimental_references"])
def test_parse_criteria(self):
crit = MPRester.parse_criteria("mp-1234 Li-*")
self.assertIn("Li-O", crit["$or"][1]["chemsys"]["$in"])
self.assertIn({"task_id": "mp-1234"}, crit["$or"])
crit = MPRester.parse_criteria("Li2*")
self.assertIn("Li2O", crit["pretty_formula"]["$in"])
self.assertIn("Li2I", crit["pretty_formula"]["$in"])
self.assertIn("CsLi2", crit["pretty_formula"]["$in"])
crit = MPRester.parse_criteria("Li-*-*")
self.assertIn("Li-Re-Ru", crit["chemsys"]["$in"])
self.assertNotIn("Li-Li", crit["chemsys"]["$in"])
comps = MPRester.parse_criteria("**O3")["pretty_formula"]["$in"]
for c in comps:
self.assertEqual(len(Composition(c)), 3, "Failed in %s" % c)
chemsys = MPRester.parse_criteria("{Fe,Mn}-O")["chemsys"]["$in"]
self.assertEqual(len(chemsys), 2)
comps = MPRester.parse_criteria("{Fe,Mn,Co}O")["pretty_formula"]["$in"]
self.assertEqual(len(comps), 3, comps)
#Let's test some invalid symbols
self.assertRaises(ValueError, MPRester.parse_criteria, "li-fe")
self.assertRaises(ValueError, MPRester.parse_criteria, "LO2")
crit = MPRester.parse_criteria("POPO2")
self.assertIn("P2O3", crit["pretty_formula"]["$in"])
colnames = ['p-type', 'n-type']
df = pandas.read_csv('Best_n_p_TEs.csv', names=colnames)
input_data_ptype = df['p-type'].tolist()[1:]
y = 0
compositions = []
for column in df:
if column == 'p-type':
te_type = 'p-type'
elif column == 'n-type':
te_type = 'n-type'
for compound in df[column].tolist()[1:]:
if not pandas.isnull(compound) and Composition(compound) not in compositions:
c = Composition(compound)
compositions.append(c)
center = get_band_center(c)
comp_results = mpr.query(c.reduced_formula,
properties=['band_gap', 'material_id', 'icsd_id', 'e_above_hull'])
try:
i = comp_results[0]
min_e_above_hull = i['e_above_hull']
except:
y += 1
print 'No compounds exists in MP for ' + c.reduced_formula
continue
for x in range(1, len(comp_results)):
if comp_results[x]['e_above_hull'] < min_e_above_hull:
i = comp_results[x]
min_e_above_hull = comp_results[x]['e_above_hull']
print ','.join([str(x) for x in
c.formula, c.reduced_formula, center + i['band_gap'] / 2, center, center, center - i[
'band_gap'] / 2, i['material_id'], i['icsd_id'], column])
csv_outfile.writerow(
from pymatgen import Composition, Element
from pymatgen.matproj.rest import MPRester
__author__ = 'Anubhav Jain'
__copyright__ = 'Copyright 2014, The Materials Project'
__version__ = '0.1'
__maintainer__ = 'Anubhav Jain'
__email__ = '*****@*****.**'
__date__ = 'Oct 01, 2014'
def get_band_center(form):
c = Composition(str(form))
prod = 1.0
for el, amt in c.get_el_amt_dict().iteritems():
prod = prod * (Element(el).X ** amt)
return -prod ** (1 / sum(c.get_el_amt_dict().values()))
if __name__ == '__main__':
mpr = MPRester()
for i in mpr.query({"elements": {"$in": ["O", "S", "Se", "Te"]}, "nelements": 2, "e_above_hull": {"$lte": 0.05}},
properties=['band_gap', 'formula', 'material_id', 'icsd_id']):
c = Composition(i['formula'])
center = get_band_center(c)
print ','.join([str(x) for x in
c.formula, c.reduced_formula, center, center + i['band_gap'] / 2, center - i['band_gap'] / 2, i[
'material_id'], i['icsd_id']])
class MPResterTest(unittest.TestCase):
def setUp(self):
self.rester = MPRester()
def test_get_data(self):
props = ["energy", "energy_per_atom", "formation_energy_per_atom",
"nsites", "unit_cell_formula", "pretty_formula", "is_hubbard",
"elements", "nelements", "e_above_hull", "hubbards",
"is_compatible", "task_ids",
"density", "icsd_ids", "total_magnetization"]
expected_vals = [-191.33812137, -6.833504334642858, -2.551358929370749,
28, {u'P': 4, u'Fe': 4, u'O': 16, u'Li': 4},
"LiFePO4", True, [u'Li', u'O', u'P', u'Fe'], 4, 0.0,
{u'Fe': 5.3, u'Li': 0.0, u'O': 0.0, u'P': 0.0}, True,
['mp-540081', 'mp-601412', 'mp-19017'],
3.4662026991351147,
[159107, 154117, 160776, 99860, 181272, 166815,
260571, 92198, 165000, 155580, 38209, 161479, 153699,
260569, 260570, 200155, 260572, 181341, 181342,
72545, 56291, 97764, 162282, 155635],
16.0002716]
for (i, prop) in enumerate(props):
if prop not in ['hubbards', 'unit_cell_formula', 'elements',
'icsd_ids', 'task_ids']:
val = self.rester.get_data("mp-19017", prop=prop)[0][prop]
self.assertAlmostEqual(expected_vals[i], val)
elif prop in ["elements", "icsd_ids", "task_ids"]:
self.assertEqual(set(expected_vals[i]),
set(self.rester.get_data("mp-19017",
prop=prop)[0][prop]))
else:
self.assertEqual(expected_vals[i],
self.rester.get_data("mp-19017",
prop=prop)[0][prop])
props = ['structure', 'initial_structure', 'final_structure', 'entry']
for prop in props:
obj = self.rester.get_data("mp-19017", prop=prop)[0][prop]
if prop.endswith("structure"):
self.assertIsInstance(obj, Structure)
elif prop == "entry":
obj = self.rester.get_data("mp-19017", prop=prop)[0][prop]
self.assertIsInstance(obj, ComputedEntry)
#Test chemsys search
data = self.rester.get_data('Fe-Li-O', prop='unit_cell_formula')
self.assertTrue(len(data) > 1)
elements = {Element("Li"), Element("Fe"), Element("O")}
for d in data:
self.assertTrue(
set(Composition(d['unit_cell_formula']).elements).issubset(
elements))
self.assertRaises(MPRestError, self.rester.get_data, "Fe2O3",
"badmethod")
def test_get_materials_id_from_task_id(self):
self.assertEqual(self.rester.get_materials_id_from_task_id(
"mp-540081"), "mp-19017")
def test_get_entries_in_chemsys(self):
syms = ["Li", "Fe", "O"]
all_entries = self.rester.get_entries_in_chemsys(syms, False)
entries = self.rester.get_entries_in_chemsys(syms)
self.assertTrue(len(entries) <= len(all_entries))
elements = set([Element(sym) for sym in syms])
for e in entries:
self.assertIsInstance(e, ComputedEntry)
self.assertTrue(set(e.composition.elements).issubset(elements))
def test_get_structure_by_material_id(self):
s1 = self.rester.get_structure_by_material_id("mp-1")
self.assertEqual(s1.formula, "Cs1")
def test_get_entry_by_material_id(self):
e = self.rester.get_entry_by_material_id("mp-19017")
self.assertIsInstance(e, ComputedEntry)
self.assertTrue(e.composition.reduced_formula, "LiFePO4")
def test_query(self):
criteria = {'elements': {'$in': ['Li', 'Na', 'K'], '$all': ['O']}}
props = ['formula', 'energy']
data = self.rester.query(criteria=criteria, properties=props)
self.assertTrue(len(data) > 6)
def test_get_exp_thermo_data(self):
data = self.rester.get_exp_thermo_data("Fe2O3")
self.assertTrue(len(data) > 0)
for d in data:
self.assertEqual(d.formula, "Fe2O3")
def test_get_dos_by_id(self):
dos = self.rester.get_dos_by_material_id("mp-2254")
self.assertIsInstance(dos, CompleteDos)
def test_get_bandstructure_by_material_id(self):
bs = self.rester.get_bandstructure_by_material_id("mp-2254")
self.assertIsInstance(bs, BandStructureSymmLine)
def test_get_structures(self):
structs = self.rester.get_structures("Mn3O4")
self.assertTrue(len(structs) > 0)
def test_get_entries(self):
entries = self.rester.get_entries("TiO2")
self.assertTrue(len(entries) > 1)
for e in entries:
self.assertEqual(e.composition.reduced_formula, "TiO2")
entries = self.rester.get_entries("TiO2", inc_structure="final")
self.assertTrue(len(entries) > 1)
for e in entries:
self.assertEqual(e.structure.composition.reduced_formula, "TiO2")
def test_get_exp_entry(self):
entry = self.rester.get_exp_entry("Fe2O3")
self.assertEqual(entry.energy, -825.5)
def test_submit_query_delete_snl(self):
s = Structure([[5, 0, 0], [0, 5, 0], [0, 0, 5]], ["Fe"], [[0, 0, 0]])
# d = self.rester.submit_snl(
# [s, s], remarks=["unittest"],
# authors="Test User <*****@*****.**>")
# self.assertEqual(len(d), 2)
# data = self.rester.query_snl({"about.remarks": "unittest"})
# self.assertEqual(len(data), 2)
# snlids = [d["_id"] for d in data]
# self.rester.delete_snl(snlids)
# data = self.rester.query_snl({"about.remarks": "unittest"})
# self.assertEqual(len(data), 0)
def test_get_stability(self):
entries = self.rester.get_entries("Fe-O")
modified_entries = []
for entry in entries:
# Create modified entries with energies that are 0.01eV higher
# than the corresponding entries.
if entry.composition.reduced_formula == "Fe2O3":
modified_entries.append(
ComputedEntry(entry.composition,
entry.uncorrected_energy + 0.01,
parameters=entry.parameters,
entry_id="mod_{}".format(entry.entry_id)))
rest_ehulls = self.rester.get_stability(modified_entries)
all_entries = entries + modified_entries
compat = MaterialsProjectCompatibility()
all_entries = compat.process_entries(all_entries)
pd = PhaseDiagram(all_entries)
a = PDAnalyzer(pd)
for e in all_entries:
if str(e.entry_id).startswith("mod"):
for d in rest_ehulls:
if d["entry_id"] == e.entry_id:
data = d
break
self.assertAlmostEqual(a.get_e_above_hull(e),
data["e_above_hull"])
def test_get_reaction(self):
rxn = self.rester.get_reaction(["Li", "O"], ["Li2O"])
self.assertIn("Li2O", rxn["Experimental_references"])
class MPDataRetrieval:
"""
MPDataRetrieval is used to retrieve data from the Materials Project database, print the results, and convert them
into an indexed/unindexed Pandas dataframe.
"""
def __init__(self, api_key=None):
"""
Args:
api_key: (str) Your Citrine API key, or None if you've set the MAPI_KEY environment variable
Returns: None
"""
api_key = api_key if api_key else os.environ["MAPI_KEY"]
self.mprester = MPRester(api_key=api_key)
def get_dataframe(self, criteria, properties, mp_decode=False, index_mpid=True):
"""
Gets data from MP in a dataframe format.
See API docs at https://materialsproject.org/wiki/index.php/The_Materials_API for more details.
Args:
criteria: (str/dict) Criteria of the query as a string or mongo-style dict. If string, it supports a
powerful but simple string criteria. E.g., "Fe2O3" means search for materials with reduced_formula
Fe2O3. Wild cards are also supported. E.g., "\*2O" means get all materials whose formula can be formed
as \*2O, e.g., Li2O, K2O, etc.
Other syntax examples:
mp-1234: Interpreted as a Materials ID.
Fe2O3 or \*2O3: Interpreted as reduced formulas.
Li-Fe-O or \*-Fe-O: Interpreted as chemical systems.
You can mix and match with spaces, which are interpreted as "OR". E.g. "mp-1234 FeO" means query for
all compounds with reduced formula FeO or with materials_id mp-1234.
Using a full dict syntax, even more powerful queries can be constructed. For example,
{"elements":{"$in":["Li", "Na", "K"], "$all": ["O"]}, "nelements":2} selects all Li, Na and K oxides.
{"band_gap": {"$gt": 1}} selects all materials with band gaps greater than 1 eV.
properties: (list) Properties to request for as a list. For example,
["formula", "formation_energy_per_atom"] returns the formula and formation energy per atom.
mp_decode: (bool) Whether to do a decoding to a Pymatgen object where possible. In some cases, it might be
useful to just get the raw python dict, i.e., set to False.
index_mpid: (bool) Whether to set the materials_id as the dataframe index.
Returns: A Pandas dataframe object
"""
if index_mpid and "material_id" not in properties:
properties.append("material_id")
data = self.mprester.query(criteria, properties, mp_decode)
df = pd.DataFrame(data, columns=properties)
if index_mpid:
df = df.set_index("material_id")
return df