Consider two particle filtering implementations: 

Implementation 1: Initialize particles by sampling from initial state distribution and assigning uniform weights. 

1. Propagate particles, retaining weights 

2. Resample according to weights 

3. Weight according to observations

Implementation 2: Initialize particles by sampling from initial state distribution. 

1. Propagate unweighted particles 

2. Weight according to observations 

3. Resample according to weights 

(i) [true or false] Implementation 2 will typically provide a better approximation of the estimated distribution than implementation 1. 

(ii) [true or false] If the transition model is deterministic then both implementations provide equally good estimates of the distribution 

(iii) [true or false] If the observation model is deterministic then both implementations provide equally good estimates of the distribution.