我正在嘗試將下面的 json 模式讀取到資料幀，我可以通過遍歷所有節點將其轉換為我喜歡的型別，但這可能需要一段時間，因為原始 json 檔案比這個示例要長得多。（以萬計）
"data": [
{
"node": {
"id": 2,
"title": "Berserk",
"main_picture": {
"medium": "https:\/\/api-cdn.myanimelist.net\/images\/manga\/1\/157897.jpg",
"large": "https:\/\/api-cdn.myanimelist.net\/images\/manga\/1\/157897l.jpg"
},
"alternative_titles": {
"synonyms": [
"Berserk: The Prototype"
],
"en": "Berserk",
"ja": "\u30d9\u30eb\u30bb\u30eb\u30af"
},
"start_date": "1989-08-25",
"end_date": "2021-09-10",
"synopsis": "Guts, a former mercenary now known as the \"Black Swordsman,\" is out for revenge. After a tumultuous childhood, he finally finds someone he respects and believes he can trust, only to have everything fall apart when this person takes away everything important to Guts for the purpose of fulfilling his own desires. Now marked for death, Guts becomes condemned to a fate in which he is relentlessly pursued by demonic beings.\n\nSetting out on a dreadful quest riddled with misfortune, Guts, armed with a massive sword and monstrous strength, will let nothing stop him, not even death itself, until he is finally able to take the head of the one who stripped him\u2014and his loved one\u2014of their humanity.\n\n[Written by MAL Rewrite]\n\nIncluded one-shot:\nVolume 14: Berserk: The Prototype",
"mean": 9.42,
"rank": 1,
"popularity": 2,
"num_list_users": 468103,
"num_scoring_users": 223269,
"nsfw": "white",
"created_at": "1970-01-01T00:00:00 00:00",
"updated_at": "2021-12-03T19:02:32 00:00",
"media_type": "manga",
"status": "on_hiatus",
"genres": [
{
"id": 1,
"name": "Action"
},
{
"id": 2,
"name": "Adventure"
},
{
"id": 10,
"name": "Fantasy"
},
{
"id": 40,
"name": "Psychological"
},
{
"id": 37,
"name": "Supernatural"
},
{
"id": 38,
"name": "Military"
},
{
"id": 6,
"name": "Demons"
},
{
"id": 8,
"name": "Drama"
},
{
"id": 14,
"name": "Horror"
},
{
"id": 41,
"name": "Seinen"
}
],
"my_list_status": {
"status": "completed",
"is_rereading": false,
"num_volumes_read": 0,
"num_chapters_read": 0,
"score": 9,
"updated_at": "2020-05-29T15:13:01 00:00"
},
"num_volumes": 41,
"num_chapters": 380,
"authors": [
{
"node": {
"id": 1868,
"first_name": "Kentarou",
"last_name": "Miura"
},
"role": "Story & Art"
}
]
},
"ranking": {
"rank": 1
}
},
{
"node": {
"id": 1706,
"title": "JoJo no Kimyou na Bouken Part 7: Steel Ball Run",
"main_picture": {
"medium": "https:\/\/api-cdn.myanimelist.net\/images\/manga\/3\/179882.jpg",
"large": "https:\/\/api-cdn.myanimelist.net\/images\/manga\/3\/179882l.jpg"
},
"alternative_titles": {
"synonyms": [
"JoJo's Bizarre Adventure Part 7: Steel Ball Run",
"SBR"
],
"en": "",
"ja": "\u30b8\u30e7\u30b8\u30e7\u306e\u5947\u5999\u306a\u5192\u967a Part7 STEEL BALL RUN"
},
"start_date": "2004-01-19",
"end_date": "2011-04-19",
"synopsis": "In the American Old West, the world's greatest race is about to begin. Thousands line up in San Diego to travel over six thousand kilometers for a chance to win the grand prize of fifty million dollars. With the era of the horse reaching its end, contestants are allowed to use any kind of vehicle they wish. Competitors will have to endure grueling conditions, traveling up to a hundred kilometers a day through uncharted wastelands. The Steel Ball Run is truly a one-of-a-kind event.\n\nThe youthful Johnny Joestar, a crippled former horse racer, has come to San Diego to watch the start of the race. There he encounters Gyro Zeppeli, a racer with two steel balls at his waist instead of a gun. Johnny witnesses Gyro using one of his steel balls to unleash a fantastical power, compelling a man to fire his gun at himself during a duel. In the midst of the action, Johnny happens to touch the steel ball and feels a power surging through his legs, allowing him to stand up for the first time in two years. Vowing to find the secret of the steel balls, Johnny decides to compete in the race, and so begins his bizarre adventure across America on the Steel Ball Run.\n\n[Written by MAL Rewrite]",
"mean": 9.25,
"rank": 2,
"popularity": 27,
"num_list_users": 178230,
"num_scoring_users": 105411,
"nsfw": "white",
"created_at": "1970-01-01T00:00:00 00:00",
"updated_at": "2021-12-03T18:56:19 00:00",
"media_type": "manga",
"status": "finished",
"genres": [
{
"id": 1,
"name": "Action"
},
{
"id": 2,
"name": "Adventure"
},
{
"id": 37,
"name": "Supernatural"
},
{
"id": 13,
"name": "Historical"
},
{
"id": 7,
"name": "Mystery"
},
{
"id": 14,
"name": "Horror"
},
{
"id": 27,
"name": "Shounen"
},
{
"id": 41,
"name": "Seinen"
}
],
"num_volumes": 24,
"num_chapters": 96,
"authors": [
{
"node": {
"id": 2619,
"first_name": "Hirohiko",
"last_name": "Araki"
},
"role": "Story & Art"
}
]
},
"ranking": {
"rank": 2
}
}
],
"paging": {
"next": "https:\/\/api.myanimelist.net\/v2\/manga\/ranking?offset=2&ranking_type=manga&limit=2&fields=id,title,main_picture,alternative_titles,start_date,end_date,synopsis,mean,rank,popularity,num_list_users,num_scoring_users,nsfw,created_at,updated_at,media_type,status,genres,my_list_status,num_volumes,num_chapters,authors{first_name,last_name},pictures,background,related_anime,related_manga,recommendations"
}
}
讓我掙扎的專欄是流派和作者。我只需要流派列中的名稱欄位，以及作者列中的 first_name、last_name 列。
我想要的格式是
genres first_name last_name
0 Action, Adventure, Fantasy, Kentarou Miura
Psychological, Supernatural ...
1 Action, Adventure, Supernatural, Hirohiko Araki
Historical, Mystery, Horror ...
我目前擁有的代碼是這個
import pandas as pd
import json
with open('./response.json') as file:
data = json.load(file)
df = pd.json_normalize(data['data'])
此代碼以這種格式提供流派和作者列
node.genres node.authors
0 [{'id': 1, 'name': 'Action'}, {'id': 2, 'name'... [{'node': {'id': 1868, 'first_name': 'Kentarou...
1 [{'id': 1, 'name': 'Action'}, {'id': 2, 'name'... [{'node': {'id': 2619, 'first_name': 'Hirohiko...
2 [{'id': 1, 'name': 'Action'}, {'id': 2, 'name'... [{'node': {'id': 1881, 'first_name': 'Eiichiro...
3 [{'id': 1, 'name': 'Action'}, {'id': 2, 'name'... [{'node': {'id': 1911, 'first_name': 'Takehiko...
4 [{'id': 8, 'name': 'Drama'}, {'id': 7, 'name':... [{'node': {'id': 1867, 'first_name': 'Naoki', ...
我想知道，如果我可以使用可以為我提供我喜歡的型別而無需遍歷所有行的技巧。
uj5u.com熱心網友回復：
或者，您可以執行以下任一操作。在有限的測驗中，在 str 方法上使用 apply() 是最快的。
選項1
df['genres'] = df.apply(lambda x: ', '.join([d['name'] for d in x['node.genres']]), axis=1)
df['first_name'] = df['node.authors'].str[0].str['node'].str['first_name']
df['last_name'] = df['node.authors'].str[0].str['node'].str['last_name']
選項 2
df['genres'] = df.apply(lambda x: ', '.join([d['name'] for d in x['node.genres']]), axis=1)
df['first_name'] = df.apply(lambda x: x['node.authors'][0]['node']['first_name'], axis=1)
df['last_name'] = df.apply(lambda x: x['node.authors'][0]['node']['last_name'], axis=1)
修改選項 2（最快）
df['genres'] = df.apply(lambda x: ', '.join([d['name'] for d in x['node.genres']]), axis=1)
df [['first_name', 'last_name']] = df.apply(lambda x: (x['node.authors'][0]['node']['first_name'], x['node.authors'][0]['node']['last_name']), axis=1)
時間結果：
answer 1 from not_speshal:
1.94 ms ± 30.1 μs per loop (mean ± std. dev. of 7 runs, 1000 loops each)
My options in order:
1.45 ms ± 45.4 μs per loop (mean ± std. dev. of 7 runs, 1000 loops each)
1.35 ms ± 37 μs per loop (mean ± std. dev. of 7 runs, 1000 loops each)
925 μs ± 6.07 μs per loop (mean ± std. dev. of 7 runs, 1000 loops each)
uj5u.com熱心網友回復：
構建作者和流派的各個 DataFrames 并加入原始資料df：
authors = pd.json_normalize(data["data"], record_path=["node",["authors"]])[["node.first_name", "node.last_name"]]
genres = pd.Series([", ".join(g["name"] for g in d) for d in df["node.genres"].tolist()]).rename("genres")
output = df.join(authors.join(genres))
轉載請註明出處，本文鏈接：https://www.uj5u.com/yidong/373178.html