Pandas 列头

索引列用这句话重置索引列df = df.reset_index()多行列就是每个父索引存在几个子索引。df.columns = [col[0] for col in df.columns]直接做成一个。

默认分类 · 04-05

Alphaflod2爬取ID号通过关键词和序列比对

import requests import pandas as pd # 读取CSV文件 esol = pd.read_csv("eSol_test.csv") # 创建一个空列表来存储entryId entry_ids = [] # 遍历每一行 for gene, sequence in zip(esol['gene'], esol['sequence']): keyname = gene res = requests.get(f"https://alphafold.ebi.ac.uk/api/search?q=%28text%3A%2a{keyname}%20OR%20text%3A{keyname}%2a%29&type=main&start=0&rows=20") found_id = None # 初始化found_id为None # 检查响应是否成功 if res.status_code == 200: for item in

默认分类 · 04-02

Python代码：多指标柱状图1

import matplotlib.pyplot as plt import numpy as np from matplotlib.patches import Rectangle from matplotlib.ticker import FixedLocator, FixedFormatter # 数据 models = ['QwQ-32B', 'DeepSeek-R1-671B', 'OpenAI-o1-mini', 'DeepSeek-R1-Distill-Llama-70B', 'DeepSeek-R1-Distill-Qwen-32B'] benchmarks = ['AIME24', 'LiveCodeBench', 'LiveBench', 'IFEval', 'BFCL'] data = [ [79.5, 79.8, 72.6, 63.6, 70.0], [63.4, 65.9, 57.5, 53.8, 57.2], [73.1, 71.6, 59.1, 57.9, 54.6], [83.9, 83.3, 84.

默认分类 · 04-02

T-SNE数据降维

以下是使用T-SNE将1280维数据降至2维或3维的分步指南：import numpy as np from sklearn.manifold import TSNE from sklearn.preprocessing import StandardScaler from sklearn.decomposition import PCA # 可选步骤 import matplotlib.pyplot as plt ### 数据预处理 标准化数据，使其均值为0，方差为1 scaler = StandardScaler() data_scaled = scaler.fit_transform(emds) tsne = TSNE( n_components=2, # 目标维度 perplexity=30, # 建议范围5-50，控制邻域大小 learning_rate=200, # 默认值，可调整范围10-1000 n_iter=1000, # 最小迭代次数，可增加至5000 random_state=42,

默认分类 · 04-02

多进程蛋白质结构PDB计算接触图

import os import numpy as np from Bio.PDB import PDBParser from multiprocessing import Pool def calculate_residue_centroid(residue): """计算氨基酸残基的几何中心（排除氢原子）""" atoms = [atom for atom in residue.get_atoms() if atom.element != 'H'] coordinates = np.array([atom.get_coord() for atom in atoms]) return coordinates.mean(axis=0) def calculate_centroid_contact_map(structure, chain_id, threshold=8.0): """生成基于质心的接触矩阵""" ch

默认分类 · 03-30
2025 不知名站点. All Rights Reserved.
Theme Jasmine by Kent Liao
新增