Я знаю, что подобные вопросы были опубликованы, но ни одно из опубликованных решений, которые я пробовал, не сработало так, как я ожидал.
У меня есть столбец во фрейме данных pandas, который является dict, и я хочу создать новый фрейм данных с каждым ключом в качестве столбца фрейма данных. Вот что содержит одна запись в talentpool_subset:
'{"Paradigms":["Agile Software Development","Scrum","DevOps","Serverless Architecture"],"Platforms":["Kubernetes","Linux","Windows","Eclipse","PagerDuty","Apache2","Docker","AWS EC2","Amazon Web Services (AWS)","Sysdig","Apache Kafka","AWS Lambda","Azure","OpenStack"],"Storage":["AWS S3","MongoDB","Cassandra","MySQL","PostgreSQL","AWS DynamoDB","Spring Data MongoDB","AWS RDS","MySQL/MariaDB","Datadog","Memcached"],"Languages":["Java","PHP","SQL","Bash","Perl","JavaScript","Python","C#","Go"],"Frameworks":["Ruby on Rails (RoR)","AWS HA",".NET","Serverless Framework","Selenium","CodeIgniter","Express.js"],"Other":["Cisco","Content Delivery Networks (CDN)","Kubernetes Operations (Kops)","Prometheus","VMware ESXi","Bash Scripting","Scrum Master","Infrastructure as Code","Performance Tuning","Serverless","System Administration","Linux System Administration","Code Review"],"Libraries/APIs":["Node.js","Jenkins Pipeline","jQuery","React","Selenium Grid"],"Tools":["Jenkins","Bitbucket","GitHub","AWS ECS","AWS IAM","Amazon CloudFront CDN","Terraform","AWS CloudFormation","Git Flow","Artifactory","Nginx","Grafana","Zabbix","Docker Compose","AWS CLI","AWS ECR","Chef","Jira","Git","Postfix","MongoDB Shell","Wowza","Amazon SQS","AWS SES","Subversion (SVN)","TeamCity","Microsoft Visual Studio","Google Kubernetes Engine (GKE)","VMware ESX","Fluentd","Sumo Logic","Slack","Apache ZooKeeper","AWS Fargate","Ansible","ELK (Elastic Stack)","Microsoft Team Foundation Server","Azure Kubernetes Service (AKS)"]}'Stack)","Microsoft Team Foundation Server","Azure Kubernetes Service (AKS)"]}'
Итак, я хотел бы, Paradigms, Platforms, Storage, Languages, et c. чтобы все были отдельными столбцами.
Я пробовал:
df = talentpool_subset.drop('skills', axis=1).join(pd.DataFrame(talentpool_subset.skills.values.tolist()))
И все же получил тот же результат:
name profile location 0
0 Hugo L. Samayoa DevOps Developer Long Beach, CA, United States {"Paradigms":["Agile Software Development","Sc...
1 Stepan Yakovenko Software Developer Novosibirsk, Novosibirsk Oblast, Russia {"Platforms":["Debian Linux","Windows","Linux"...
2 Slobodan Gajic Software Developer Sremska Mitrovica, Vojvodina, Serbia {"Platforms":["Firebase","XAMPP"],"Storage":["...
3 Bruno Furtado Montes Oliveira Visual Studio Team Services (VSTS) Developer Niterói - State of Rio de Janeiro, Brazil {"Paradigms":["Agile","CQRS","Azure DevOps"],"...
4 Jennifer Aquino Query Optimization Developer West Ryde, New South Wales, Australia {"Paradigms":["Automation","ETL Implementation...
Любые идеи о том, как решить эту проблему ?