rietx · compare settings

Does this correction actually make the fit better — and where?

Standard

Variants

Reference for Δχ²

Export

Cumulative Δχ² vs the reference

Σ2θ'≤2θ(δ²variant − δ²ref). Falling = winning, rising = losing, flat = doing nothing. The slope says where the change acted, which is what separates a correction that models physics from one that absorbs it. The endpoint is the whole χ² difference.

Nothing drawn yet: tick the variants to compare and press Run.

Weighted residual (yobs − ycalc)/σ

Same vertical scale for every variant, so smaller is literally better.

Observed and calculated

Statistics

Read the diagnostics below before believing a Rwp improvement. Some variants here cannot change Rwp by construction — capillary absorption is an exact reparameterisation of the scale and Biso — so the parameter columns are where their effect shows up.

Diagnostics

Refined parameters