Quick Start

Installation and loadingOne-loop matching
Defining the UV theoryMatching onto the Warsaw basis
Tree-level matchingNext steps
Matchete is a Mathematica package for the automated matching of UV theories onto effective field theories (EFTs) at tree level and one-loop order, using functional methods. This tech note walks through a complete example: the Standard Model (SM) extended by a heavy real scalar singlet ϕ. We integrate out ϕ at tree level and at one-loop order, simplify the resulting EFT Lagrangian to an on-shell basis, and finally express the result in terms of the Wilson coefficients of the Warsaw basis of the SMEFT. More details on the framework can be found in arXiv:2212.04510.

Main functions for a matching computation with Matchete:

LoadModel
loads a predefined model and returns its Lagrangian
DefineField
defines a new field
DefineCoupling
defines a new coupling
FreeLag
returns the kinetic and mass terms of fields
Match
integrates out the heavy fields and returns the EFT Lagrangian
GreensSimplify
reduces an EFT Lagrangian to a Green's basis
EOMSimplify
reduces an EFT Lagrangian to an on-shell basis using field redefinitions
MapEffectiveCouplings
maps between different EFT bases by expressing the couplings of a target Lagrangian in terms of the couplings of an input Lagrangian (e.g. obtained from the matching)
NiceForm
prints expressions in a human-readable form
Installation and loading
Matchete can be installed by evaluating the following command, which downloads the package and installs it in the Applications folder of Mathematica's base directory:
Afterwards, Matchete is loaded with:
Matchete can be updated to the latest release by running:
All of the following inputs are evaluated in a single Matchete session.
Defining the UV theory
Matchete comes with a number of predefined models (see GetModels). We start by loading the SM with LoadModel, which defines the SM gauge groups, fields, and couplings, and returns the SM Lagrangian:
Load the SM Lagrangian:
Next, we define the new heavy scalar with DefineField. The option Mass->{Heavy,M} specifies that ϕ is a heavy field with mass M, which is integrated out when matching. The option SelfConjugate->True makes ϕ a real field:
Define the heavy scalar singlet:
All couplings appearing in the Lagrangian have to be defined with DefineCoupling. The option NiceForm specifies how a coupling is displayed by NiceForm; here, the couplings μϕ and λϕ are printed with subscripts:
Fields and couplings are entered with the short-hand notation label[indices], where the indices are given in the order specified upon definition. Repeated indices are summed over. The kinetic and mass terms of ϕ are obtained with FreeLag.
Write the new-physics Lagrangian, including all renormalizable interactions of ϕ with the Higgs doublet H:
The full UV Lagrangian is the sum of both. CheckLagrangian performs a series of consistency checks, e.g. of hermiticity and gauge invariance:
The same model is also available as a predefined model and can be loaded with LoadModel["Singlet_Scalar_Extension"]. How to write your own model files is explained in Model Files.
Tree-level matching
Match integrates out all heavy fields. The option EFTOrder specifies the maximal mass dimension of the EFT operators and LoopOrder the order in the loop expansion (0 for tree level, 1 for tree level plus one loop, and {1} for the pure one-loop piece).
Match the UV Lagrangian onto the EFT at tree level up to dimension six:
Show the terms that differ from the SM Lagrangian (HcSimplify combines terms with their Hermitian conjugates):
The matching result is not given in a minimal basis. GreensSimplify reduces it to a Green's basis using integration by parts and other operator identities, while EOMSimplify additionally removes redundant operators with field redefinitions, yielding an on-shell basis.
Reduce the tree-level result to an on-shell basis:
One-loop matching
With LoopOrder->1, Match computes the tree-level and one-loop contributions. One-loop terms are multiplied by hbar, which stands for a loop factor , and depend on the renormalization scale through logarithms of (μbar2). Poles in ϵ, where 𝒹=4-2ϵ, are related to the difference of UV divergences between the UV model and the EFT and are removed by the appropriate counterterms; for the matching conditions in the scheme they can be dropped.
Match at one-loop order and drop the 1/ϵ poles:
Reduce the result to an on-shell basis:
To keep the result compact, EOMSimplify may introduce effective couplings, e.g. for the shifted Higgs mass parameter. They can be shown with PrintEffectiveCouplings and replaced by their definitions with ReplaceEffectiveCouplings.
Show the effective couplings introduced in the simplification:
Individual operator classes can be extracted with SelectOperatorClass. For instance, the contributions to the operator with six Higgs fields and no derivatives are:
Matching onto the Warsaw basis
Finally, we express the matching result in terms of the Wilson coefficients of the SMEFT in the Warsaw basis. Matchete provides the SMEFT Lagrangian as a predefined model and MapEffectiveCouplings determines the coefficients of the target Lagrangian from the matching result.
Load the dimension-six SMEFT Lagrangian in the Warsaw basis:
Compute the matching conditions for the SMEFT couplings. Since MapEffectiveCouplings simplifies both Lagrangians with EOMSimplify, the matching result can be passed directly:
For instance, the matching condition for the Wilson coefficient CH□ is:
Note that couplings appearing on both sides of the matching conditions, such as the Higgs quartic coupling λ, refer to the SMEFT on the left-hand side and to the UV theory on the right-hand side.
Next steps
The matching results can be further processed, e.g. exported to the WCxf format with ExportWCxf for the running and phenomenological analysis with other tools, or used to derive Feynman rules with FeynmanRules and ExportUFO; see BSM to EFT: Universal Feynman Output. More details on the internal representation and the available functionality are given in the other tech notes and in the Matchete guide.
When publishing results obtained with Matchete, please cite the papers suggested by SuggestBibliography for the functionality used in your session: