Skip to content
Skills Library
Science & research · Analysis

quickfit

/quickfit

Use when fitting a RooWorkspace dataset with quickFit, generating an Asimov dataset with quickAsimov, computing asymptotic CLs limits with quickLimit, generating toy datasets with quickToy, rebinning datasets with quickRebin, or calculating nuisance parameter pulls with quickPull.

Repository
marketplace ↗
Licence
MIT
Last repository change
16.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
3
Deprecation

No signals detected. Repository changed on 16.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when fitting a RooWorkspace dataset with quickFit, generating an Asimov dataset with quickAsimov, computing asymptotic CLs limits with quickLimit, generating toy datasets with quickToy, rebinning datasets with quickRebin, or calculating nuisance parameter pulls with quickPull.

Original author description

Use when fitting a RooWorkspace dataset with quickFit, generating an Asimov dataset with quickAsimov, computing asymptotic CLs limits with quickLimit, generating toy datasets with quickToy, rebinning datasets with quickRebin, or calculating nuisance parameter pulls with quickPull.

How it works

  1. Overview
  2. Standalone Installation
  3. quickFit — Fitting a Workspace
  4. quickAsimov — Generating Asimov Datasets
  5. quickLimit — Asymptotic CLs Limits
  6. quickToy — Generating Toy Datasets
  7. quickRebin — Rebinning a Dataset
  8. quickPull — NP Pull to a POI
  9. Gotchas
  10. Interop
Original SKILL.md headings · 193 lines, 899 words
SKILL.md preview
---
name: quickfit
description: >-
  Use when fitting a RooWorkspace dataset with quickFit, generating an Asimov
  dataset with quickAsimov, computing asymptotic CLs limits with quickLimit,
  generating toy datasets with quickToy, rebinning datasets with quickRebin, or
  calculating nuisance parameter pulls with quickPull.
---

# quickFit

## Overview

quickFit is a suite of command-line tools for statistical operations on a
RooWorkspace. It provides rapid fitting, Asimov generation, CLs limit
computation, toy generation, dataset rebinning, and NP pull calculations. All
tools share a common interface and depend on ROOT and
[RooFitExtensions](https://gitlab.cern.ch/atlas_higgs_combination/software/RooFitExtensions).

quickFit is included in the StatAnalysis release (all executables available
after `asetup StatAnalysis,0.7,latest`). It can also be built standalone.

Repository: https://gitlab.cern.ch/atlas_higgs_combination/software/quickFit

## Standalone Installation

```bash
git clone ssh://git@gitlab.cern.ch:7999/atlas_higgs_combination/software/quickFit.git
cd quickFit
source setup_lxplus.sh   # sets up ROOT, cmake, boost from CVMFS
sh scripts/install_roofitext.sh   # install RooFitExtensions if not present
mkdir build && cd build
cmake ..
make && make install
cd ..
```

After the build, re-run `source setup_lxplus.sh` before using executables.

## quickFit — Fitting a Workspace

```bash
# Basic fit with floating POIs
quickFit -f ws.root -d dataset -p mu=1_-5_5

# Float mu_ggH, fix mu_VBF=2
quickFit -f ws.root -d dataset -p mu_ggH=1_-5_5,mu_VBF=2

# Fix all NPs matching ATLAS_* prefix
quickFit -f ws.root -d dataset -p mu=1_-5_5 -n "ATLAS_*"

# Run HESSE and MINOS on top of MIGRAD
quickFit -f ws.root -d dataset -p mu=1_-5_5 --hesse 1 --minos 1

# Save fit results to file
quickFit -f ws.root -d dataset -p mu=1_-5_5 -o output.root --savefitresult 1

# Parallel evaluation (multi-core or GPU)
quickFit --useModularL=true --numCPU=4 -f ws.root -d dataset -p mu=1_-5_5
quickFit --EvalBackend=cuda -f ws.root -d dataset -p mu=1_-5_5
```

### Parameter syntax (`-p`)

| Form                          | Meaning                                        |
| ----------------------------- | ---------------------------------------------- |
| `mu=1_-5_5`                   | Float `mu` with initial value 1, range [−5, 5] |
| `mu=2`                        | Fix `mu` at 2                                  |
| `mu_ggH=1_-5_5,mu_VBF=1_-5_5` | Multiple params, comma-separated               |

### Key quickFit options

| Option            | Description                               |
| ----------------- | ----------------------------------------- |
| `-f`              | Input workspace ROOT file                 |
| `-w`              | Workspace name (default: `combWS`)        |
| `-m`              | ModelConfig name (default: `ModelConfig`) |
| `-d`              | Dataset name to fit                       |
| `-p`              | POI/parameter configuration               |
| `-n`              | Fix NPs matching pattern                  |
| `-o`              | Output ROOT file                          |
| `--savefitresult` | 0 = NLL only, 1 = full RooFitResult       |
| `--hesse`         | Run HESSE after MIGRAD (1=yes)            |
| `--minos`         | Run MINOS (1=yes)                         |
| `--useModularL`   | Enable modular likelihood evaluation      |
| `--numCPU`        | Number of cores for parallel evaluation   |
| `--EvalBackend`   | Evaluation backend (`cpu`, `cuda`, etc.)  |
| `--minTolerance`  | Minimizer tolerance                       |

## quickAsimov — Generating Asimov Datasets

`quickAsimov` reads an XML card describing fit and Asimov generation actions.

```bash
quickAsimov -x card.xml -w combWS -m ModelConfig -d obsData
```

### XML card format

```xml
<!DOCTYPE Asimov SYSTEM 'asimovUtil.dtd'>
<Asimov InputFile="input.root" OutputFile="output.root" POI="mu">
  <Action Name="Prepare" Setup="" Action="nominalNuis:nominalGlobs"/>
  <Action Name="Fit" Setup="mu=1_0_5"
          Action="fit:matchglob:savesnapshot"
          SnapshotNuis="conditionalNuis_1"
          SnapshotGlob="conditionalGlob_1"/>
  <Action Name="asimovData_1" Setup="mu=1"
          Action="genasimov:nominalGlobs"/>
</Asimov>
```

### Action keywords (colon-separated in `Action` attribute)

| Keyword           | Meaning                                                  |
| ----------------- | -------------------------------------------------------- |
| `fit`             | Perform maximum likelihood fit                           |
| `genasimov`       | Generate Asimov dataset (once per action list)           |
| `savesnapshot`    | Save parameter snapshot (once per action list)           |
| `matchglob`       | Match global observables to NP values (use with `reset`) |
| `reset`           | Reset parameters to state before current action list     |
| `raw`             | Reset to state before any actions                        |
| `fixsyst`         | Fix all constrained NPs                                  |
| `float`           | Float NPs fixed by `fixsyst` or `Setup`                  |
| `<snapshot name>` | Load a saved snapshot                                    |

## quickLimit — Asymptotic CLs Limits

```bash
quickLimit -f ws.root -w combWS -m ModelConfig -d obsData -p mu -o limits.root
```

- Output: ROOT file with limit histograms + `.txt` file with numeric results.
- Only one POI is processed for limit setting; use the first listed with `-p`.
- Other POIs can still be configured via `-p` for fixing/floating.

## quickToy — Generating Toy Datasets

Requires snapshots already saved in the workspace (e.g. from `quickAsimov`).

```bash
quickToy -f ws.root -d toyData \
  -p mu=1,mu_ggH=1 \
  -s conditionalGlob_1,conditionalNuis_1 \
  -o output.root --seed 1
```

Use a different `--seed` for each independent toy.

## quickRebin — Rebinning a Dataset

Converts an unbinned dataset to a pseudo-binned one for faster fitting:

```bash
quickRebin -f output.root -d toyData -o output_binned.root -r 500
```

Creates `toyDatabinned` in the output file with 500 bins. Categories where the
number of entries is less than the bin count remain unbinned.

## quickPull — NP Pull to a POI

```bash
quickPull -f ws.root -w combWS -d combData \
  -p mu=1_0_5,mu_VBF=1 \
  --nui_param ATLAS_EG_RESOLUTION_ALL \
  --pull_POI mu \
  -o ATLAS_EG_RESOLUTION_ALL.root
```

## Gotchas

- **Custom classes**: If the workspace uses custom RooFit classes (from
  `RooFitExtensions` or local code), ensure `LD_LIBRARY_PATH` includes the
  library directory before opening the workspace, or the program will crash.
- **`matchglob` requires `reset`**: Always pair `matchglob` with `reset` at the
  end of the same action list to restore global observable values.
- **`genasimov` once per action list**: Including it more than once causes an
  error.
- **Workspace name defaults**: All tools default to `combWS`/`ModelConfig`;
  specify `-w` and `-m` explicitly if names differ.
- **ROOT ≥ 6.18 required**: Older ROOT versions are not supported.

## Interop

- **workspaceCombiner**: Produces combined workspaces that quickFit operates on.
- **xmlAnaWSBuilder / HistFactory**: Produces input workspaces; quickFit is the
  recommended fitting tool for these.
- **StatAnalysis**: All quickFit executables are available after
  `asetup StatAnalysis,0.7,latest`.
- **xRooFit**: Alternative Python/C++ API for fitting and limit setting built
  into ROOT within StatAnalysis.

## Support

Contact: quickFit-user@cern.ch

## Docs

https://gitlab.cern.ch/atlas_higgs_combination/software/quickFit

Versions

The database has no other skills named quickfit.

Database snapshot: 13.11.2026 · package version 11ca853a99f0 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100toy
    通过 Toy CLI 完成 B站 Toy 平台的创作者操作(发布 / 更新 / 预览 / 查询 / 视频绑定 / 排查)。触发词:Toy 发布、发布 Toy、toy create、上传 Toy、Toy 更新、toy update、改 Toy 封面、改 Toy 密码、Toy 预览、Toy 提交审核、我的 Toy、toy mylist、Toy 列表、Toy 统计、toy stats、Toy PV/UV、Toy 登录、toy login、Toy 发布记录、toy history、toy cli、Toy 白屏、Toy 404、Toy 页面打不开、Toy…
  • 100dataset-discovery
    Multi-source ML dataset discovery. Search HuggingFace Hub, OpenML, GitHub, and paper cross-references for datasets relevant to a research task. Use when asked to "find datasets for", "search ML datasets", "what datasets exist for", or "discover training data for".
  • 100xroofit
    Use when building RooFit statistical models with xRooFit, constructing workspaces with channels and samples, adding systematics or nuisance parameters, fitting models to data, constructing NLL functions, diagnosing fit convergence failures (status codes, covariance quality),…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.