Comment on 'He's frequency formulation for nonlinear oscillators'

Geng and Cai suggested a modified He's frequency-amplitude formulation for nonlinear oscillators ( Geng and Cai 2007 Eur. J. Phys. 28 923-31). We point out the accuracy of both the original one and the modified version depends upon the chosen location point, and so far we have no general rule for the location choice. The method of weighted residuals is suggested to overcome the shortcoming.