For completely randomised design for k treatments and n observations, $$y_{if}$$ = response from the $$j^{th}$$ unit receiving $$i^{th}$$ treatment, $$\sum_i \sum_j y_{ij} = y_{..}$$ and $$\sum_j^{ni} y_{ij} = y_{i.}$$. Which of the following options is correct?

$$\sum_{i=1}^{k} \sum_{j=1}^{n_{i}}\left(y_{ij} - \overline{y}_{i}\right)^{2}$$ represents sum of squares of due to error

$$\sum_{i=1}^{k}\left(\overline{y}_{i}- \overline{y}_{..}\right)^{2}$$ represents sum of squares due to treatments

$$\sum_{i=1}^{k} \sum_{j=1}^{n_{i}}\left(y_{ij} - \overline{y}_{..}\right)^{2}$$ represents total sum of squares

$$\sum_{i=1}^{k} n_{i} \left(\overline{y}_{i.}- \overline{y}_{..}\right)^{2}$$ represents sum of squares due to treatments