anndata k折交叉

如何将anndata拆分为k份

import scanpy as sc
import anndata as ad

class KSplitAnndata:
    @staticmethod
    def _base_split(data: ad.AnnData, k: int) -> list:
        adata = data.copy()
        num = adata.n_obs // k
        adata_list = []
        for i in range(k):
            if num <= adata.n_obs:
                adata_list.append(sc.pp.subsample(adata, n_obs=num, copy=True))
                adata = adata[~adata.obs_names.isin(adata_list[i].obs_names)]
            else:
                adata_list.append(adata)
        return adata_list
    
    @staticmethod
    def k_split(data: ad.AnnData, k: int, batch_key: str = None) -> list:
        """
        
        :param data: anndata object
        :param k: k fold
        :param batch_key: split by batch, default is None
        """
        adata = data.copy()
        if batch_key:
            adata_batch_list = []
            adata_list = []
            batch_tuple = set(adata.obs[batch_key])
            for batch in batch_tuple:
                adata_batch = adata[adata.obs[batch_key]==batch]
                adata_batch_list.append(KSplitAnndata._base_split(adata_batch, k))
            for i in range(k):
                adata_list.append(ad.concat([adata_batch[i] for adata_batch in adata_batch_list]))
        else:
            adata_list = KSplitAnndata._base_split(adata, k)
        return adata_list

# example
adata_list = KSplitAnndata.k_split(adata, 5, "batch")
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值