In Rudin's functional analysis , let $X$ denote a topological vector space . He states that the open sets of $X$ are precisely those that are unions of translates of members of the local base of $0$ .
Since for any nonempt subset $E$ , we can find some $x\in E$ , then we have $$E=(-x+E)+x$$ and $-x+E$ is a neighborhood of $0$ , so it suffice to show every neighborhood of $0$ can be write as a union of the local base of $0$ . However , for each neighborhood of $0$ , by definition of local base , we can only find an element of the local base which is a subset of the neighborhood , so how can we get the desired conclusion ?