代码如下：

from Bio import Entrez
import time
import pandas as pd
import matplotlib.pyplot as plt
from concurrent.futures import ThreadPoolExecutor, as_completed

# 设置你的邮箱（NCBI 要求）
Entrez.email = "your_email@example.com"
Entrez.tool = "PubMedYearStatsTool"

def search_pubmed(keyword, retmax=10000):
    print(f"Searching PubMed for '{keyword}'...")
    handle = Entrez.esearch(db="pubmed", term=keyword, retmax=retmax, datetype="pdat")
    record = Entrez.read(handle)
    id_list = record["IdList"]
    print(f"Found {len(id_list)} articles.")
    return id_list

def fetch_summary_batch(article_ids):
    """获取一个批次的摘要信息"""
    try:
        handle = Entrez.esummary(db="pubmed", id=",".join(article_ids), retmode="xml")
        summaries = Entrez.read(handle)
        time.sleep(0.5)  # 控制请求频率
        return summaries
    except Exception as e:
        print(f"Error fetching batch: {e}")
        return []

def fetch_all_summaries(id_list, batch_size=200):
    """并发获取所有摘要信息"""
    batches = [id_list[i:i + batch_size] for i in range(0, len(id_list), batch_size)]
    all_summaries = []

    with ThreadPoolExecutor(max_workers=4) as executor:  # 最多并发 4 个线程
        future_to_batch = {executor.submit(fetch_summary_batch, batch): batch for batch in batches}
        for future in as_completed(future_to_batch):
            summaries = future.result()
            if summaries:
                all_summaries.extend(summaries)

    return all_summaries

def extract_years(summaries):
    year_count = {}
    for summary in summaries:
        try:
            pub_date = summary.get("PubDate", "")
            if pub_date:
                year = pub_date.split(" ")[0][:4]  # 提取前四位作为年份
                if year.isdigit() and len(year) == 4:
                    year_int = int(year)
                    if 2015 <= year_int <= 2025:  # 只保留 2015 - 2025 的数据
                        year_count[year] = year_count.get(year, 0) + 1
        except Exception as e:
            continue
    return year_count

def plot_yearly_trend(year_count, keyword):
    df = pd.DataFrame.from_dict(year_count, orient='index', columns=['Count'])
    df.index = df.index.astype(int)
    df.sort_index(inplace=True)

    # 过滤 2015 - 2025
    df = df[(df.index >= 2015) & (df.index <= 2025)]

    plt.figure(figsize=(7, 6))
    plt.bar(df.index, df['Count'], color='skyblue')
    plt.title(f"Number of PubMed Articles on '{keyword}' by Year (2015–2025)")
    plt.xlabel('Year')
    plt.ylabel('Number of Articles')
    plt.xticks(range(2015, 2026))
    plt.grid(True, linestyle='--', alpha=0.5)
    plt.tight_layout()
    plt.show()

def main():
    keyword = input("Enter your search keyword: ")
    id_list = search_pubmed(keyword, retmax=10000)
    print("Fetching metadata (summary)...")
    summaries = fetch_all_summaries(id_list)
    print("Extracting publication years (2015–2025)...")
    year_count = extract_years(summaries)
    
    print("\nArticles per year (2015–2025):")
    for year in sorted(year_count):
        print(f"{year}: {year_count[year]}")

    plot_yearly_trend(year_count, keyword)

if __name__ == '__main__':
    main()

结果：

手动绘制：

import pandas as pd
import matplotlib.pyplot as plt
# 数据
# data = {
#     'Year': [2025, 2024, 2023, 2022, 2021, 2020],
#     'Count': [1198, 2297, 1571, 962, 733, 502]
# }

data = {
    'Year': [2025, 2024, 2023, 2022, 2021, 2020, 2019, 2018, 2017, 2016],
    'Count': [3261, 7026, 5379, 4339, 4298, 3043, 2128, 1665, 1342, 1183]
}

# 创建 DataFrame
df = pd.DataFrame(data)
# 绘制柱状图
plt.figure(figsize=(7, 6))
bars = plt.bar(df['Year'], df['Count'], color='skyblue')
# 设置标题和坐标轴标签
keyword="single cell transcriptomics"
plt.title(f"Number of PubMed Articles on '{keyword}' by Year (2016–2025)")
plt.xlabel('Year')
plt.ylabel('Count')
# 显示网格线
plt.grid(axis='y', linestyle='--', alpha=0.7)
# 确保所有年份都显示出来
plt.xticks(df['Year'], rotation=45)  # 横轴标签倾斜角度为45度，防止重叠
# 显示数值在柱子上方
for bar in bars:
    yval = bar.get_height()
    plt.text(bar.get_x() + bar.get_width()/2, yval + 10, int(yval), ha='center')
# 自动调整布局，防止标签被截断
plt.tight_layout()
# 显示图表
plt.show()