def test_retrieve_authors_from_sourceyear():
make_database(test_cache, drop=True)
conn = connect_database(test_cache)
# Variables
expected_sources = [22900]
expected_years = [2005, 2010]
df = pd.DataFrame(product(expected_sources, expected_years),
columns=["source_id", "year"],
dtype="int64")
# Populate cache
expected = query_pubs_by_sourceyear(expected_sources,
expected_years[0],
refresh=refresh)
expected["source_id"] = expected["source_id"].astype(np.int64)
expected["afid"] = expected["afid"].astype(int).astype(str)
expected = expected.sort_values(["auids", "afid"]).reset_index(drop=True)
expected = expected[['source_id', 'year', 'auids', 'afid']]
expected["auids"] = expected["auids"].str.split(";")
insert_data(expected, conn, table="sources_afids")
# Retrieve from cache
incache, missing = retrieve_authors_from_sourceyear(df, conn)
incache["afid"] = incache["afid"].astype(int).astype(str)
incache = incache.sort_values(["auids", "afid"]).reset_index(drop=True)
assert_frame_equal(incache, expected)
assert_frame_equal(missing, df.tail(1).reset_index(drop=True))
def test_retrieve_author_info_authoryear():
make_database(test_cache, drop=True)
conn = connect_database(test_cache)
# Variables
table = "author_year"
expected_auth = [53164702100, 57197093438]
search_auth = [55317901900]
year = 2016
df2 = pd.DataFrame(expected_auth + search_auth,
columns=["auth_id"],
dtype="int64")
df2["year"] = year
# Insert data
fill = robust_join(expected_auth, sep=') OR AU-ID(')
q = f"(AU-ID({fill})) AND PUBYEAR BEF {year+1}"
d = build_dict(ScopusSearch(q, refresh=refresh).results, expected_auth)
expected = pd.DataFrame.from_dict(d, orient="index", dtype="int64")
expected = expected.sort_index().rename_axis('auth_id').reset_index()
expected["year"] = year
expected = expected[[
'auth_id', 'year', 'first_year', 'n_pubs', 'n_coauth'
]]
insert_data(expected, conn, table=table)
# Retrieve data
incache, missing = retrieve_author_info(df2, conn, table)
assert_frame_equal(incache, expected)
assert_equal(missing['auth_id'].tolist(), search_auth)
assert_equal(missing['year'].tolist(), [year])
def test_retrieve_authors():
make_database(test_cache, drop=True)
conn = connect_database(test_cache)
# Variables
expected_auth = [53164702100, 57197093438]
df = pd.DataFrame(expected_auth, columns=["auth_id"], dtype="int64")
expected_cols = [
'auth_id', 'eid', 'surname', 'initials', 'givenname', 'affiliation',
'documents', 'affiliation_id', 'city', 'country', 'areas'
]
# Retrieve data
incache, missing = retrieve_authors(df, conn)
assert_equal(incache.shape[0], 0)
assert_equal(incache.columns.to_list(), expected_cols)
assert_equal(missing, expected_auth)
def test_retrieve_author_info_authorncits():
make_database(test_cache, drop=True)
conn = connect_database(test_cache)
# Variables
table = "author_ncits"
data = {
"auth_id": [53164702100, 53164702100],
"year": [2010, 2017],
"n_cits": [0, 6]
}
expected = pd.DataFrame(data, dtype="int64")
# Insert data
insert_data(expected, conn, table=table)
# Retrieve data
cols = ["auth_id", "year"]
incache, tosearch = retrieve_author_info(expected[cols], conn, table)
assert_frame_equal(incache, expected)
assert_true(tosearch.empty)
def test_retrieve_authors_insert():
make_database(test_cache, drop=True)
conn = connect_database(test_cache)
# Variables
expected_auth = [53164702100, 57197093438]
search_auth = [55317901900]
expected_cols = [
'auth_id', 'eid', 'surname', 'initials', 'givenname', 'affiliation',
'documents', 'affiliation_id', 'city', 'country', 'areas'
]
# Insert data
q = f"AU-ID({robust_join(expected_auth, sep=') OR AU-ID(')})"
res = pd.DataFrame(AuthorSearch(q, refresh=refresh).authors, dtype="int64")
res["auth_id"] = res["eid"].str.split("-").str[-1]
res = res[expected_cols]
insert_data(res, conn, table="authors")
# Retrieve data
df = pd.DataFrame(expected_auth + search_auth,
columns=["auth_id"],
dtype="int64")
incache, missing = retrieve_authors(df, conn)
assert_equal(incache.shape[0], 2)
assert_equal(missing, [55317901900])
def __init__(self,
identifier,
year,
refresh=False,
period=None,
eids=None,
sql_fname=None):
"""Class to represent a scientist.
Parameters
----------
identifier : list of int
List of Scopus Author IDs of the scientist.
year : str or numeric
Year for which characteristics should be defined for.
refresh : boolean or int (optional, default=False)
Whether to refresh cached results (if they exist) or not. If int
is passed, results will be refreshed if they are older than
that value in number of days.
eids : list (optional, default=None)
A list of scopus EIDs of the publications of the scientist. If
it is provided, the scientist's properties are set based on these
publications, instead of the list of publications obtained from
the Scopus Author ID(s).
period: int (optional, default=None)
In additional starting x years prior to the treatment year,
which is also used to compute characteristics in the treatment
year.
sql_fname : str (optional, default=None)
The path of the SQLite database to connect to. If None, will use
the path specified in config.ini.
Raises
------
Exception
When there are no publications for the author until the
provided year.
"""
self.identifier = identifier
self.year = int(year)
if not sql_fname:
sql_fname = config.get('Filepaths', 'Database')
self.sql_conn = connect_database(sql_fname)
# Read mapping of fields to sources
df, names = read_fields_sources_list()
self.field_source = df
self.source_names = names.set_index("source_id")["title"].to_dict()
# Load list of publications
if eids:
q = f"EID({' OR '.join(eids)})"
else:
q = f"AU-ID({') OR AU-ID('.join([str(i) for i in identifier])})"
integrity_fields = ["eid", "author_ids", "coverDate", "source_id"]
res = base_query("docs", q, refresh, fields=integrity_fields)
self._publications = [p for p in res if int(p.coverDate[:4]) <= year]
if not len(self._publications):
text = "No publications found for author "\
f"{'-'.join([str(i) for i in identifier])} until {year}"
raise Exception(text)
self._eids = eids or [p.eid for p in self._publications]
# First year of publication
pub_years = [p.coverDate[:4] for p in self._publications]
self._first_year = int(min(pub_years))
self._period_year = self.year - (period or (self.year + 1)) + 1
if self._period_year < self._first_year:
self._period_year = 0
# Count of citations
search_ids = eids or identifier
self._citations = count_citations(search_ids, self.year + 1,
identifier)
# Coauthors
self._coauthors = set(extract_authors(
self._publications)) - set(identifier)
# Period counts simply set to total if period is or goes back to None
if self._period_year:
pubs = [
p for p in self._publications
if self._period_year <= int(p.coverDate[:4]) <= year
]
self._publications_period = pubs
if not len(self._publications_period):
text = "No publications found for author "\
f"{'-'.join([str(i) for i in identifier])} until "\
f"{year} in a {self._period_year}-years period"
raise Exception(text)
eids_period = [p.eid for p in self._publications_period]
n_cits = count_citations(eids_period, self.year + 1, identifier)
self._citations_period = n_cits
self._coauthors_period = set(
extract_authors(self._publications_period))
self._coauthors_period -= set(identifier)
else:
self._coauthors_period = None
self._publications_period = None
self._citations_period = None
# Author search information
source_ids = set(
[int(p.source_id) for p in self._publications if p.source_id])
self._sources = add_source_names(source_ids, self.source_names)
self._active_year = int(max(pub_years))
mask = df["source_id"].isin(source_ids)
self._fields = df[mask]["asjc"].astype(int).tolist()
self._main_field = get_main_field(self._fields)
if not self._main_field[0]:
text = "Not possible to determine research field(s) of "\
"researcher. Functionality is reduced."
warn(text, UserWarning)
# Most recent geolocation
afid = find_main_affiliation(identifier, self._publications, year)
self._affiliation_id = afid
try:
aff = AffiliationRetrieval(afid, refresh=refresh)
self._affiliation_country = aff.country
self._affiliation_name = aff.affiliation_name
self._affiliation_type = aff.org_type
except (Scopus404Error, ValueError):
self._affiliation_country = None
self._affiliation_name = None
self._affiliation_type = None
self._language = None
# Author name from profile with most documents
df = get_authors(self.identifier,
self.sql_conn,
refresh=refresh,
verbose=False)
au = df.sort_values("documents", ascending=False).iloc[0]
self._subjects = [a.split(" ")[0] for a in au.areas.split("; ")]
self._surname = au.surname or None
self._first_name = au.givenname or None
name = ", ".join([self._surname or "", au.givenname or ""])
if name == ", ":
name = None
self._name = name
# -*- coding: utf-8 -*-
"""Tests for processing.getting module."""
from os.path import expanduser
from nose.tools import assert_equal, assert_true
from pandas import DataFrame
from sosia.establishing import connect_database
from sosia.processing.getting import get_authors
test_cache = expanduser("~/.sosia/test.sqlite")
test_conn = connect_database(test_cache)
refresh = 30
def test_query_authors():
auth_list = [6701809842, 55208373700]
auth_data = get_authors(auth_list, test_conn, refresh=refresh)
assert_true(isinstance(auth_data, DataFrame))
expected_cols = ["auth_id", "eid", "surname", "initials", "givenname",
"affiliation", "documents", "affiliation_id", "city",
"country", "areas"]
assert_equal(auth_data.columns.tolist(), expected_cols)
assert_equal(auth_data["auth_id"].tolist(), auth_list)
assert_equal(auth_data["surname"].tolist(), ["Harhoff", "Baruffaldi"])
