# Difference between revisions of "Analysis results for configurations 30-33"

Line 1: | Line 1: | ||

== Summary == | == Summary == | ||

− | This posting gives results for an ILC based analysis | + | This posting gives results for an ILC based analysis for configurations 30-33 introduced in the posting [[Simulations for configurations with different frequency coverage]]. The main goal is to understand the effect of frequency coverage on the bias in r. Results are given for foreground models 01-03, which are PySM models a1d1f1s1, a2d4f1s3, a2d7f1s3, and model 06, which is a model based on MHD simulations. Results for all XX.YY combinations are based on 1000 simulations with a fiducial value for r of zero. The seed of the random number generator is chosen so noise realizations for the different configurations are independent. This results in a statistical uncertainty for the bias of just above 1e-5. |

+ | The tables summarize the results, more details are provided below. For the PySM models, the bias decreases as additional frequency bands are added. For the MHD model the bias is small for all configurations. | ||

+ | |||

=== Results for foreground model 01 === | === Results for foreground model 01 === | ||

Line 47: | Line 49: | ||

|- | |- | ||

|} | |} | ||

− | |||

== Foreground model 01 == | == Foreground model 01 == |

## Revision as of 22:46, 31 March 2019

## Contents

## Summary

This posting gives results for an ILC based analysis for configurations 30-33 introduced in the posting Simulations for configurations with different frequency coverage. The main goal is to understand the effect of frequency coverage on the bias in r. Results are given for foreground models 01-03, which are PySM models a1d1f1s1, a2d4f1s3, a2d7f1s3, and model 06, which is a model based on MHD simulations. Results for all XX.YY combinations are based on 1000 simulations with a fiducial value for r of zero. The seed of the random number generator is chosen so noise realizations for the different configurations are independent. This results in a statistical uncertainty for the bias of just above 1e-5.

The tables summarize the results, more details are provided below. For the PySM models, the bias decreases as additional frequency bands are added. For the MHD model the bias is small for all configurations.

### Results for foreground model 01

Configuration | Bias on r before marginalization | Bias on r after marginalization |
---|---|---|

30 | 2.4e-4 | 2.0e-4 |

31 | 2.7e-4 | 2.4e-4 |

32 | 2.0e-4 | 1.5e-4 |

33 | 1.3e-4 | 0.9e-4 |

### Results for foreground model 03

Configuration | Bias on r before marginalization | Bias on r after marginalization |
---|---|---|

30 | 2.5e-4 | 2.1e-4 |

31 | 2.5e-4 | 2.2e-4 |

32 | 2.0e-4 | 1.4e-4 |

33 | 1.5e-4 | 1.1e-4 |

### Results for foreground model 06

Configuration | Bias on r before marginalization | Bias on r after marginalization |
---|---|---|

30 | 0.4e-4 | -0.1e-4 |

31 | 0.2e-4 | -0.2e-4 |

32 | 0.7e-4 | 0.0e-4 |

33 | -0.3e-4 | -0.7e-4 |

## Foreground model 01

Configuration | Bias on r before marginalization | Bias on r after marginalization |
---|---|---|

30 | 2.4e-4 | 2.0e-4 |

31 | 2.7e-4 | 2.4e-4 |

32 | 2.0e-4 | 1.5e-4 |

33 | 1.3e-4 | 0.9e-4 |

### Histograms before marginalization

### Histograms after marginalization

### ILC spectra and residuals for realizations 0000-0003

#### Configuration 30

#### Configuration 31

#### Configuration 32

#### Configuration 33

## Foreground model 02

Configuration | Bias on r before marginalization | Bias on r after marginalization (*) |
---|---|---|

30 | 2.5e-4 | 2.0e-4 |

31 | 2.3e-4 | 2.4e-4 |

32 | 1.9e-4 | 1.5e-4 |

33 | 1.4e-4 | 0.9e-4 |

### Histograms before marginalization

### Histograms after marginalization

### ILC spectra and residuals for realizations 0000-0003

#### Configuration 30

#### Configuration 31

#### Configuration 32

#### Configuration 33

## Foreground model 03

Configuration | Bias on r before marginalization | Bias on r after marginalization |
---|---|---|

30 | 2.5e-4 | 2.1e-4 |

31 | 2.5e-4 | 2.2e-4 |

32 | 2.0e-4 | 1.4e-4 |

33 | 1.5e-4 | 1.1e-4 |

### Histograms before marginalization

### Histograms after marginalization

### ILC spectra and residuals for realizations 0000-0003

#### Configuration 30

#### Configuration 31

#### Configuration 32

#### Configuration 33

## Foreground model 06

Configuration | Bias on r before marginalization | Bias on r after marginalization |
---|---|---|

30 | 0.4e-4 | -0.1e-4 |

31 | 0.2e-4 | -0.2e-4 |

32 | 0.7e-4 | 0.0e-4 |

33 | -0.3e-4 | -0.7e-4 |