About the arc from 'test.fits'

[WentaoLuo]

Is there a lens galaxy or the smoothed profile of the dark matter that creates the arcs?

[WentaoLuo]

I will upload the python code soon. it is based on Rachel's work SHERA.

[WentaoLuo]

The simulated Gaussian has bigger dispersion.

[WentaoLuo]

These noise map are simulated with noise variance=58.75, typical SDSS noise variance.

[WentaoLuo]

The unit is count^2.

[WentaoLuo]

The poisson part may be wrong. I will check and modify it later.

[linan7788626]

Hi @WentaoLuo , I just merged your SDSS images , and will be working on it soon. thanks a lot.

[WentaoLuo]

I will focus on galaxy next.

[linan7788626]

so you think the noise model is right?

[WentaoLuo]

Yes, I talked with Rachel. it is right.

[linan7788626]

Hi wentao, I have merged your code and update the script for producing lensed images. If the noise model is correct, then, so far, the only thing we need is the images of lenses, and please update your commits of the code in the same way. Thanks. BTW, Good luck to you about your paper. :-)

[WentaoLuo]

Simulated de Vaucouleurs with R_e=2.9'', miu_e=27.989 z=0.3 with gaussian noise

[WentaoLuo]

This one is a figure without noise in ln(image) plot. It seems it is too big. Maybe we have to add sky level rather than noise only.

[WentaoLuo]

./lens_galaxy.py --E 0.5 --Phi 45 --R_e 2.0 --zl 0.1 --NoiseVar 59

Once I put the lens_galaxy.py to this repo, everyone can simulate an de Vaucouleurs image by adding parameters like ellipticity --E, position angle --Phi or effective radius --R_e e.t.c

[WentaoLuo]

The velocity dispersion of a galaxy with R_e and miu_e, can be calculated also in lens_galaxy.py.

[WentaoLuo]

Yes, we definitely need to add sky to mimic SDSS survey.

[WentaoLuo]

How much information we lost from the sky!!!

[WentaoLuo]

This galaxy has miu_e~ 18.8825

[WentaoLuo]

I mean effective brightness

[linan7788626]

it looks great, I will get into it later. thanks. did you make a pull request? if so I will merge it asap.

[WentaoLuo]

this one is with Poisson noise. I think Guassian is better.

[WentaoLuo]

I will modify it and check 1, the fundamental plane relation 2, noise modelling

And then put it on git and send a pull request.

[linan7788626]

OK, thanks, I will read your new scripts later. By the way, I did not see any noise in the latest figure. is the scales of noise and flux right?

On Dec 16, 2015, at 3:54 PM, Wentao Luo notifications@github.com wrote:

I will modify it and check 1, the fundamental plane relation 2, noise modelling

And then put it on git and send a pull request.

— Reply to this email directly or view it on GitHub https://github.com/linan7788626/Arc_Finding_with_CMU/issues/2#issuecomment-165255119.

[WentaoLuo]

The Poisson noise model has much smaller variance than Gaussian one, that is the reason. I use exactly the model by Rachel. For now, I think we should use Gaussian, because it seems more realistic.

[linan7788626]

can I ask why the poisson noise is much smaller than Gaussian? Is it realistic?

[linan7788626]

Hi Wentao, Could you please make a pull request for your new code ? so that I can merge them in the master branch. Thanks

[WentaoLuo]

I did a major modification of the simulation code. The image magnitude will be given by velocity dispersion, effective radius(arcsec), redshift. The noise from Rachel works well after this modification.

I will show you some example with fix redshift z=0.5, effetive radius=10, and velocity dispersion=100,200,300

[WentaoLuo]

[WentaoLuo]

Next, we fix vd=300, radius_eff=10, z=0.2,0.3,0.5

[WentaoLuo]

[WentaoLuo]

Sorry the size is fixed

[WentaoLuo]

Let's fix zl=0.3, vd=300 and R_eff=5,8,10

[WentaoLuo]

[WentaoLuo]

I will push the code soon.

[WentaoLuo]

$./lens_galaxy.py --NoiseVar 59 --Phi 45 --E 0.3 --R_eff 8 --zl 0.3 --VelDis 300

The basic command is like that. I will write a simple readme and then push the code and file soon after I finish the readme.

[WentaoLuo]

After this modification, the difference between the two noise models are not as huge as before. There was a bug when calculating the counts of SDSS detectors.

[WentaoLuo]

[WentaoLuo]

[WentaoLuo]

Poisson has smaller standard deviation than Gaussian

[linan7788626]

Oh, it looks great, this is the noise we need, thanks a lot.

On Dec 18, 2015, at 2:05 PM, Wentao Luo notifications@github.com wrote:

https://cloud.githubusercontent.com/assets/6350744/11906191/cbc531fe-a598-11e5-95a8-bc6285483d73.png — Reply to this email directly or view it on GitHub https://github.com/linan7788626/Arc_Finding_with_CMU/issues/2#issuecomment-165886245.

[WentaoLuo]

Debugging is tiring. Haha

[WentaoLuo]

Larger velocity dispersion, the brighter. If we set vd=300 km/s. The noise will be not as obvious as we set vd=150 or 100.

[linan7788626]

haha, BTW, how about the bug of your 2d kappa map? If needed, we can talk on it later.

On Dec 18, 2015, at 2:07 PM, Wentao Luo notifications@github.com wrote:

Debugging is tiring. Haha

— Reply to this email directly or view it on GitHub https://github.com/linan7788626/Arc_Finding_with_CMU/issues/2#issuecomment-165887022.

[WentaoLuo]

Sure, I am focusing on this big bug. Later I will take care of that one. I could not focus at two things at the same time. Haha

[WentaoLuo]

I think we should further to modify our pipeline a little bit to generate rgb color images of galaxy. like the above. This one is without noise. We should follow Lupton et al 2004 MNRAS 116 to do this. I will work on it.

[linan7788626]

Hi Wentao,

It looks great, I am still on my way back to Chicago, I will working on this during the weekend. thanks a lot. Happy new year! :)

-Nan

On Dec 30, 2015, at 2:02 PM, Wentao Luo notifications@github.com wrote:

https://cloud.githubusercontent.com/assets/6350744/12056593/13714896-af06-11e5-90a3-e5965351a514.png I think we should further to modify our pipeline a little bit to generate rgb color images of galaxy. like the above. This one is without noise. We should follow Lupton et al 2004 MNRAS 116 to do this. I will work on it.

— Reply to this email directly or view it on GitHub https://github.com/linan7788626/Arc_Finding_with_CMU/issues/2#issuecomment-168065062.

[linan7788626]

Hi Wentao,

How could I convert the unit of flux of the lens galaxy from number counts to magnitude properly? Say, I have a 2d map as shown in the figure your attached, there should be some basic relation between number counts and magnitude, right?

Thanks, Nan

On Dec 30, 2015, at 10:40 PM, LiNan linan7788626@gmail.com wrote:

Hi Wentao,

It looks great, I am still on my way back to Chicago, I will working on this during the weekend. thanks a lot. Happy new year! :)

-Nan

On Dec 30, 2015, at 2:02 PM, Wentao Luo <notifications@github.com mailto:notifications@github.com> wrote:

https://cloud.githubusercontent.com/assets/6350744/12056593/13714896-af06-11e5-90a3-e5965351a514.png I think we should further to modify our pipeline a little bit to generate rgb color images of galaxy. like the above. This one is without noise. We should follow Lupton et al 2004 MNRAS 116 to do this. I will work on it.

— Reply to this email directly or view it on GitHub https://github.com/linan7788626/Arc_Finding_with_CMU/issues/2#issuecomment-168065062.

[WentaoLuo]

[linan7788626]

thanks wentao, it looks great, BTW, what is the angular resolution of this image?

On Jan 18, 2016, at 11:17 AM, Wentao Luo notifications@github.com wrote:

https://cloud.githubusercontent.com/assets/6350744/12397800/712db56a-bddd-11e5-906e-5f4e2e2e7057.png — Reply to this email directly or view it on GitHub https://github.com/linan7788626/Arc_Finding_with_CMU/issues/2#issuecomment-172595737.

[WentaoLuo]

This is an example image from acs at z~1.0. Sorry for the late update, I took my kid to get vaccine last Thurday and Friday was my older daughter's birthday.

[WentaoLuo]

Angular resolution interms of arcseconds?