• methods: data analysis;
  • cosmic background radiation


A new method to constrain the local non-linear coupling parameter inline image based on a fast wavelet decomposition is presented. Using a multiresolution wavelet adapted to the healpix pixelization, we have developed a method that is ∼102 times faster than previous estimators based on isotropic wavelets and ∼103 faster than the KSW bispectrum estimator, at the resolution of the Wilkinson Microwave Anisotropy Probe (WMAP) data. The method has been applied to the WMAP 7-yr V+W combined map, imposing constraints on inline image of inline image at the 95 per cent CL. This result has been obtained after correcting for the contribution of the residual point sources which has been estimated to be inline image. In addition, a Gaussianity analysis of the data has been carried out using the third-order moments of the wavelet coefficients, finding consistency with Gaussianity. Although the constraints imposed on inline image are less stringent than those found with optimal estimators, we believe that a very fast method, as the one proposed in this work, can be very useful, especially bearing in mind the large amount of data that will be provided by future experiments, such as the Planck satellite. Moreover, the localization of wavelets allows one to carry out analyses on different regions of the sky. As an application, we have separately analysed the two hemispheres defined by the dipolar modulation proposed by Hoftuft et al. We do not find any significant asymmetry regarding the estimated value of inline image in those hemispheres.