Supporting dataset for "Non-Gaussianity in D3-brane inflation", arXiv:2105.03637

Licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

This deposit is a supporting dataset that accompanies the paper "Non-Gaussianity in D3-brane inflation" (arXiv:2105.03637). It contains the primary trajectory catalogue used to generate the statistical results reported in this paper. It also contains a precise specification of the inflationary model (kinetic mixing matrix and potential) in the form of a CppTransport .model file, and supporting files needed to build the COSMOSIS pipeline used to compute observables.

Please note that the .model file is in a non-standard format. If you wish to process it, you will require a specific build of CppTransport (commit hash 35c5ad8f).

This deposit contains the following files:

Main trajectory catalogue

• catalogue.csv. This contains all parameters needed to specify the potential for a particular trajectory. (Initial conditions are not required, because each trajectory begins at the same field-space location. These initial conditions are specified in the COSMOSIS configuration file d3brane_deltamax_38_all_values.ini.) It also contains values for the summary inflationary observables. The columns contained in this file are listed below.

CppTransport setup files

As explained above, the .model file uses a custom format associated with the interface to COSMOSIS. A public release of this interface is planned, but the final .model file format used may change. These files should always be used with CppTransport commit hash 35c5ad8f.

• d3brane_deltamax_38_all.model. CppTransport .model file that fully specifies the D3-brane kinetic mixing matrix and potential. 
• christoffel.txt. Contains pre-computed expressions for the components of the connexion. This file is automatically ingested by the CppTransport translator in 35c5ad8f. It should be placed in the same folder as the .model file.
• inv_metric.txt. Contains pre-computed expressions for the inverse field-space metric (kinetic mixing matrix). This file is automatically ingested by the CppTransport translator in 35c5ad8f. It should be placed in the same folder as the .model file.
• riemann.txt. Contains pre-computed expressions for the Riemann tensor associated with the field-space metric. This file is automatically ingested by the CppTransport translator in 35c5ad8f. It should be placed in the same folder as the .model file.

PyTransport setup files

• DBraneSetup.py. PyTransport setup file to install the D3-brane model.

COSMOSIS configuration files

These are used to specify the COSMOSIS pipeline used to compute observables.

• d3brane_deltamax_38_all_mcmc.ini. Main COSMOSIS configuration file specifying the pipeline.
• d3brane_deltamax_38_all_priors.ini. Specifies priors for each sample parameter. Used only during catalogue construction.
• d3brane_deltamax_38_all_values.ini. Specifies fixed parameters of the potential (\(T_3\), \(a_0\), \(\phi_{\mathrm{UV}}\)), cosmological parameters (\(\Omega_c h^2\), \(\Omega_b h^2\), \(h\), \(\tau\)), and field-space initial conditions.

COMOSIS module files

• class_interface.py. A modified version of the CLASS interface to COSMOSIS that accepts a sampled power spectrum (written to a separate file) rather than specifying the power spectrum using \(A_s\),  \(n_s\). If the COSMOSIS configuration files are not modified, the pipeline will expect to find this interface in the location

  <COSMOSIS directory>/modules/CLASS_full_Pk/class_interface.py

  The other parts of the CLASS module can be copied to this folder. Only class_interface.py needs to be replaced.

Fields included in trajectory catalogue

The main trajectory catalogue catalogue.csv contains a large number of fields.

• trajectory. Unique trajectory label, beginning at 1. Numeric labels used in the corresponding science paper refer to this identified.
• Q, alpha. See Table 1 of the science paper.
• ReBlm_xxx, ImBlm_xxx. Real and imaginary parts of the complex Wilson coefficient for "non-real" zero modes (i.e. modes with at least one of \(R\), \(m_1\), \(m_2\) not zero) appearing in the contributions to the potential from the conifold zero modes. Normalized as in Eq. (2.30) of arXiv v1 of the science paper. Drawn from a Gaussian distribution with mean zero and standard deviation unity. The xxx label is associated with an internal ordering of the modes.
• ReDlm_xxx, ImDlm_xxx. Real and imaginary parts of the complex Wilson coefficient for "non-real" modes appearing in the contributions to the potential from the inhomogeneous term \(|\Lambda^2|\) in the field equation for \(\Phi_-\). Normalized as in the discussion below Eq. (2.42) of arXiv v1 of the science paper. Drawn from a Gaussian distribution with mean zero and standard deviation unity. The xxx label is associated with an internal ordering of the modes.
• A_lm. Real Wilson coefficient for real modes (i.e. those with \(R = m_1 = m_2 = 0\)) appearing in the contributions to the potential from the conifold zero modes. Normalized as in Eq. (2.30) of arXiv v1 of the science paper. Drawn from a Gaussian distribution with mean zero and standard deviation unity. The xxx label is associated with an internal ordering of the modes.
• C_lm. Real Wilson coefficient for real modes appearing in the contributions to the potential from the inhomogeneous term \(|\Lambda^2|\) in the field equation for \(\Phi_-\). Normalized as in the discussion below Eq. (2.42) of arXiv v1 of the science paper. Drawn from a Gaussian distribution with mean zero and standard deviation unity. The xxx label is associated with an internal ordering of the modes.
• As, At, r. Values of the power spectra at \(k = 0.002 / \mathrm{Mpc}\). Defined as in Step 1 in §3.1.1 of arXiv v1 of the science paper.
• ns, nt. Values of the spectral indices \(n_s\) and \(n_t\) at \(k = 0.002 / \mathrm{Mpc}\). Obtained by performing a fit as described in §3.1.2 of arXiv v1 of the science paper. Notice that nt is not used for the tensor spectral index estimates reported in the paper.
• nsfull, ntfull. Fits for the spectral indices using the full wavenumber range provided to CLASS. Has no clear meaning when the spectrum is not accurately fit by a power law over this range of wavenumbers. Used mostly for comparison with ns and nt.
• epsilon, eta. Values for the slow-roll parameters \(\epsilon = - \dot{H}/H^2\) and \(\eta = \mathrm{d}\ln \epsilon / \mathrm{d}N\) sampled 60 e-folds before the end of inflation. The value of epsilon is used to estimate \(n_t\) as described in the science paper.
• kpiv. Wavenumber of the pivot scale \(k = 0.002 / \text{Mpc}\) in CppTransport internal units normalized so that \(k=1\) is the mode that exits the horizon 15 e-folds after the initial point.
• Npiv. Horizon-exit time of the pivot scale \(k = 0.002/\text{Mpc}\) measured from \(N=0\) at the end of inflation.
• Nefolds. Total number of e-folds attained in this model, measured from \(N=0\) at the initial point.
• normmassmatrixeigenvalueN-M. Values of \(m_N^2 / H^2\), where \(m_N^2\) is the \(N^{\text{th}}\) ordered eigenvalue of the mass-squared matrix (from light to heavy). \(M=1\) is 55 e-folds from the end of inflation; \(M=2\) is 2.5 e-folds from the end of inflation; \(M=3\) is 1 e-fold from the end of inflation; \(M=4\) corresponds to the end of inflation.
• Bequi, fNLequi. Bispectrum shape function and reduced bispectrum (respectively) on the equilateral configuration described in Step 2, §3.1.1 of arXiv v1 of the science paper.
• Bfold, fNLfold. Bispectrum shape function and reduced bispectrum (respectively) on the folded configuration described in Step 2, §3.1.1 of arXiv v1 of the science paper.
• planck2015. Log-likelihood computed by the planck COSMOSIS module using Planck2015 data products.
• ClassInput. Set to 1 if a full CMB power spectrum was computed, otherwise 0.
• ErrStatus. Set to 0 if no errors occurred during computation.
• CppTTime. Total compute time for all observables, measured in seconds.