用组内相同的值填充缺失值(匹配已知的IMEI和丢失的手机模型)

I have a time series/panel data including encrypted IMEI (unique per cellphone but not unique per user) and cellphone brands. Not too many missing values exist in the IMEI but lots of missing values in cellphone brand. Since the unique IMEI corresponds to a unique phone, we can infer the missing brands by IMEI. My solution is to construct a IMEI to brand table and I have a time series/panel data including encry