Understanding Document Evaluation Metrics in Data Analysis

Hm - not sure why adding this column: retrieved_documents_eval["score"] allowed all the other values to be calculated?

ndcg - takes into account the document ordering, and their scores - my documents didnt come from OpenSearch with any scores (have to fix this)
precision - "what % of retrieved docs were relevant"? A simple label count
hit rate - "what % of rows have at least 1 relevant doc?"

Is this correct? Feel like I'm missing something