Fig. 7From: Assessing walking ability using a robotic gait trainer: opportunities and limitations of assist-as-needed control in spinal cord injuryLeft: actual vs predicted TUG: the unregularized linear model is created with the variable selected with Bolasso, K knee TS. The error bars show the confidence interval of the predicted observations. In blue, the identity line is shown. Right: the predictor K knee TS vs the predicted TUG (black line). CI (continuous red line) and PI (dashed red line) are calculated from the second round of bootstrapping. The reciprocal of the TUG data in seconds are used for the model, to make it consistent with the 10MWT data, which are expressed in m/sBack to article page