pyEuropePMC

ORCID and NIH iCite clients

OrcidClient reads researcher profiles and works from the ORCID public API, and ICiteClient reads citation metrics for PubMed articles from NIH iCite. Both are enrichment clients that look up known identifiers; they are not multi-source search sources. Neither needs authentication.

from pyeuropepmc.features.enrich import ICiteClient, OrcidClient

ORCID

OrcidClient(rate_limit_delay=1.0, timeout=15, cache_config=None) reads public records from the ORCID public API v3.0 (https://pub.orcid.org/v3.0).

Profiles

from pyeuropepmc.features.enrich import OrcidClient

with OrcidClient() as client:
    profile = client.get_profile("0000-0002-1825-0097")

if profile:
    print(profile["name"])
    print(f"{len(profile['works'])} works")
    for work in profile["works"][:5]:
        print(f"  {work['year']}: {work['title']} (DOI: {work['doi']})")

get_profile(orcid) returns a dict, or None when the iD cannot be read or the record cannot be retrieved. enrich(identifier) returns the same dict.

Key	Type	Content
`name`	`str`	Given and family names
`given_name`, `family_name`	`str`
`credit_name`	`str` or `None`	Published name
`other_names`, `keywords`	`list[str]`
`biography`	`str` or `None`
`urls`	`dict[str, str]`	Link name to URL
`employments`, `educations`	`list[dict]`	Dicts with `organization`, `department`, `role`, `start` and `end`, read from the record’s `affiliation-group` entries. `start` and `end` are years as strings; `end` is `""` for a current position
`works`	`list[dict]`	Works as returned by `get_works()`; repeated DOIs are dropped

Works

from pyeuropepmc.features.enrich import OrcidClient

with OrcidClient() as client:
    works = client.get_works("https://orcid.org/0000-0002-1825-0097")

for work in works:
    print(work["year"], work["type"], work["title"])

get_works(orcid) returns a list of dicts with title, doi (normalized, or None), year (int or None), type, journal_title (str or None), visibility and path, or an empty list when the iD cannot be read or the request fails.

Values the ORCID record leaves empty (null in the API response), such as a missing biography or publication date, come back as None or an empty list.

Identifier formats

0000-0002-1825-0097, https://orcid.org/0000-0002-1825-0097 and orcid.org/0000-0002-1825-0097 are all accepted; the first iD found in the string is used.

NIH iCite

ICiteClient(rate_limit_delay=0.5, timeout=15) reads metrics from the iCite API (https://icite.od.nih.gov/api).

from pyeuropepmc.features.enrich import ICiteClient

with ICiteClient() as client:
    metrics = client.enrich_many(["32791984", "32882182"])
    single = client.enrich("32791984")

for pmid, values in metrics.items():
    print(f"PMID {pmid}: {values['citation_count']} citations, "
          f"RCR {values['rcr']}, NIH percentile {values['percentile']}")

Method	Returns
`enrich(identifier)`	Metrics dict for one PMID, or `None`
`enrich_many(pmids)`	`dict[str, dict]` keyed by PMID, from one request; PMIDs iCite does not know are absent
`get_rcr(pmid)`, `get_percentile(pmid)`, `get_citation_count(pmid)`	One value or `None`; each call makes a request

Each metrics dict has these keys:

Key	Type	Content
`pmid`	`str`
`rcr`	`float` or `None`	Relative Citation Ratio
`percentile`, `nih_percentile`	`float` or `None`	NIH percentile of the RCR
`citation_count`	`int` or `None`	Citations counted by iCite
`citations_per_year`	`float` or `None`
`expected_citations`	`float` or `None`	Expected citations per year for the article’s field
`field_citation_ratio`	`float` or `None`
`is_research_article`, `provisional`	`bool`
`year`	`int` or `None`	Publication year

ICiteMetrics model

The client returns dicts. pyeuropepmc.models.ICiteMetrics is a separate Pydantic model with the fields pmid, rcr, percentile, citation_count, expected_citations, field_citation_ratio, nih_percentile and is_controversial; every field except pmid defaults to 0, 0.0 or False. It accepts extra keys without validating them, so use these exact names:

from pyeuropepmc.models import ICiteMetrics

metrics = ICiteMetrics(pmid="32791984", citation_count=42, rcr=1.5, percentile=85.0, expected_citations=28.0)
print(metrics.rcr, metrics.field_citation_ratio)  # 1.5 0.0