## PointingModelOffsets » History » Version 14

Version 13
*(Lucas Beaufore, 07/12/2013 12:56 PM)*
→
Version 14/29
*(Lucas Beaufore, 07/12/2013 01:28 PM)*

h1. Pointing Model Offsets

In order to study the effects the pointing model may have on the offsets, new offsets were calculated with possible corrections from the effects of the hexapod trim values.

The following histograms show the distribution of the offsets before trim corrections, and with a trim correction of both addition and subtraction.

RA offset with: No correction, addition of a scaled (.0176 arcsec/micron) trim_dy value, subtraction of the scaled trim_dy value

|!{width:100%}dec_standard.png! |!{width:100%}dec_corrected_y.png! | |!{width:100%}dec_corrected_minus_y.png! |

Dec offset with: No correction, addition of a scaled (.0176 arcsec/micron) trim_dx value, subtraction of the scaled trim_dx value

|!{width:100%}ra_standard.png! |!{width:100%}ra_corrected_x.png! | |!{width:100%}ra_corrected_minus_x.png! |

The effect of the corrections are arguably more visible on a 3D plot. The following show the tweak values and exposure ID values (essentially time, in order to better show the exposure groups)

versus the offsets.

Blue = Original offsets

Green = Original offsets + (.0176 * trim)

Red = Original offsets - (.0176 * trim)

Declination offsets vs trim_dy (vs expid):

|!{width:100%}dec_all_corrections_back.png! | !{width:100%}dec_all_corrections_side.png! |

RA offsets vs trim_dx (vs expid):

|!{width:70%}ra_all_corrections_side.png! |

By applying trim corrections, the offsets should no longer include the part of the offset due to hexapod corrections.

As can be seen, adding the correction (green) makes each set of four data points taken in the same position much more horizontal, removing correlation with trim values.

Subtracting the correction makes each set of four data point correlate much more strongly with trim values.

After applying the correction, the offsets can then be reexamined for correlations with the telescopes position. The plots below illustrate the relations. Each set of graphs

has the correction applied in each of the directions.

The following two plots show the Declination offsets plotted against the declination of the telescope. The color scale shows the HA of the telescope.

No correction, correction Correction added and correction subtracted, respectively

|!{width:100%}dec_teldec_ha.png! |!{width:100%}dec_corrected_teldec_ha.png! |!{width:100%}dec_neg_corrected_teldec_ha.png! |

The following two plots show the RA offsets plotted against the declination of the telescope. The color scale shows the HA of the telescope.

No correction, correction Correction added and correction subtracted, respectively

|!{width:100%}ra_corrected_teldec_ha.png! |!{width:100%}ra_neg_corrected_teldec_ha.png! |

p3. Interesting observations

By adding the correction, the range of offsets decreases, and the correlation of the offsets with trim values is mostly removed. However, the offsets are on average FARTHER from 0

By subtracting the correction, the range of offsets increases, and the correlation of the offsets with trim values is mostly remvoed. However, the offsets are on average CLOSER to 0.

In order to study the effects the pointing model may have on the offsets, new offsets were calculated with possible corrections from the effects of the hexapod trim values.

The following histograms show the distribution of the offsets before trim corrections, and with a trim correction of both addition and subtraction.

RA offset with: No correction, addition of a scaled (.0176 arcsec/micron) trim_dy value, subtraction of the scaled trim_dy value

|!{width:100%}dec_standard.png! |!{width:100%}dec_corrected_y.png! | |!{width:100%}dec_corrected_minus_y.png! |

Dec offset with: No correction, addition of a scaled (.0176 arcsec/micron) trim_dx value, subtraction of the scaled trim_dx value

|!{width:100%}ra_standard.png! |!{width:100%}ra_corrected_x.png! | |!{width:100%}ra_corrected_minus_x.png! |

The effect of the corrections are arguably more visible on a 3D plot. The following show the tweak values and exposure ID values (essentially time, in order to better show the exposure groups)

versus the offsets.

Blue = Original offsets

Green = Original offsets + (.0176 * trim)

Red = Original offsets - (.0176 * trim)

Declination offsets vs trim_dy (vs expid):

|!{width:100%}dec_all_corrections_back.png! | !{width:100%}dec_all_corrections_side.png! |

RA offsets vs trim_dx (vs expid):

|!{width:70%}ra_all_corrections_side.png! |

By applying trim corrections, the offsets should no longer include the part of the offset due to hexapod corrections.

As can be seen, adding the correction (green) makes each set of four data points taken in the same position much more horizontal, removing correlation with trim values.

Subtracting the correction makes each set of four data point correlate much more strongly with trim values.

After applying the correction, the offsets can then be reexamined for correlations with the telescopes position. The plots below illustrate the relations. Each set of graphs

has the correction applied in each of the directions.

The following two plots show the Declination offsets plotted against the declination of the telescope. The color scale shows the HA of the telescope.

No correction, correction Correction added and correction subtracted, respectively

|!{width:100%}dec_teldec_ha.png! |!{width:100%}dec_corrected_teldec_ha.png! |!{width:100%}dec_neg_corrected_teldec_ha.png! |

The following two plots show the RA offsets plotted against the declination of the telescope. The color scale shows the HA of the telescope.

No correction, correction Correction added and correction subtracted, respectively

|!{width:100%}ra_corrected_teldec_ha.png! |!{width:100%}ra_neg_corrected_teldec_ha.png! |

p3. Interesting observations

By adding the correction, the range of offsets decreases, and the correlation of the offsets with trim values is mostly removed. However, the offsets are on average FARTHER from 0

By subtracting the correction, the range of offsets increases, and the correlation of the offsets with trim values is mostly remvoed. However, the offsets are on average CLOSER to 0.