Issue/437/magnification bias #438

marina-ricci · 2021-10-01T16:30:11Z

No description provided.

…gration may still be optimized

feat: notebooks to reproduce the figures of the v1.0 paper * add Paper_v1.0.ipynb * update Paper_v1.0.ipynb * update examples/Paper_v1.0.ipynb * add support/paper_formating.py * make Paper_v1.0.ipynb use support/paper_formating.py * improve plots on examples/Paper_v1.0.ipynb * fix comment on Paper nb * select case for paper * improve plot * rm unused lines * update text in Paper nb * reset Example1 to master version * update to new ver of code * update gt plot (add gx) * put paper files in one dir * New notebook to reproduce Fig4 of the v1.0 paper * Normalize pdf in the corner plot - need to have the same number of samples for both cases * Update version on Fig3 to answer Referee 3''s comment of v1.0 paper * update Fig3_update with different fits * rm old Fig3 * renamed: Fig3_update.ipynb -> Fig3.ipynb * update text on Fig3 * Make nicer mcmc plot for paper * rename paper notebooks * update text in gt_and_use_case.ipynb * Moving new notebooks for v1.0 paper to the proper directory * Small change to NB to solve issue foreground galaxies * Fix small bug in mock_data * Small adjustment to figure * Small adjustment to figure * Change legend * Small update to notebook to fit only the bins with sufficient number of galaxies * Fix bug with SRD z distribution * Update gt_and_use_case with galaxy density instead of galaxy number * Add comment to NB * Typo in NB * update text and format in gt_and_use_case.ipynb * update plot * Update backend comparison plot to start at higher radii * Revert to previous range and shade up to 5.e-2 Mpc * update backend validation tests with optimised CCL integration for the v1.0 paper * Removed dots from abbreviations (ana. -> ana and opt. -> opt). Just to have one notation and simplify a bit the legend. * Update __init__.py Update version number Co-authored-by: m-aguena <aguena@if.usp.br> Co-authored-by: Michel Aguena <aguena@lapp.in2p3.fr> Co-authored-by: Mariana Penna Lima <pennalima@gmail.com>

coveralls · 2021-10-01T16:37:16Z

Coverage remained the same at 100.0% when pulling db63826 on issue/437/magnification_bias into d97a61f on main.

marina-ricci · 2021-10-25T09:54:27Z

This issue is also linked to issue #444.

…7/magnification_bias

marina-ricci · 2021-10-25T15:44:37Z

This issue covers the first step to integrate weak lensing magnification observables in CLMM. Here I added the "magnification bias" observable in the theory module.

include magnification bias in theory
write corresponding unit test
write corresponding examples in notebooks

m-aguena · 2021-10-27T08:42:12Z

@marina-ricci it looks like the error is on the value of the magnification bias test now.

combet · 2021-10-29T10:15:36Z

clmm/theory/generic.py

+    magnification_bias_from_magnification = np.array(
+        magnification)**(np.array([alpha]).T - 1)


@marina-ricci - This gives a warning RuntimeWarning: invalid value encountered in power whenever attempting to take non-integer power of negative values (of the magnification). This happened when running the notebook for alpha=-0.5, and the corresponding magnification bias is then nan. It's not necessarily an issue, but I'm wondering if this can be made clearer in a comment in the code or warning for the user? But maybe that might be more confusing than helpful... It just got me confused for a bit.

I am not sure what to do. The magnification can only be negative in the strong lensing regime so in practice this is not so problematic. But I'm not sure what should be the exact formula here (returning a complex number, a nan..). I'll investigate.

For negative magnifications, the output will be nan by default.

Right - we're not going to make use of a complex number, so I think a nan is fine. I'm more thinking in the line of having an explicit warning for the user, like the one we have if some sources are in front of the cluster...

An explicit warning will be helpful for users. I was confused by the RuntimeWarning.

Ok, I'll add a simple warning and maybe we can think of a more general one to warn the user when we enter the strong lensing regime.

@combet , @hsinfan1996 , Ok, it should be done now.

m-aguena

Can you make sure the lines are shorter than 100 characters? You can check in the files you changed with the command grep -x '.\{100,\}' my_file.py

m-aguena · 2021-11-04T16:12:50Z

clmm/theory/func_layer.py

@@ -9,12 +9,12 @@
 import numpy as np

 from . import generic
-from . generic import compute_reduced_shear_from_convergence
+from . generic import *


please do not import *

m-aguena · 2021-11-04T16:13:34Z

clmm/theory/generic.py

-    convergence : array_like
-        Convergence
+    shear : array_like, float
+            Shear


there is an extra tab here

also on lines 23 and 28

m-aguena · 2021-11-04T16:14:20Z

clmm/theory/parent_class.py

@@ -281,7 +283,7 @@ def eval_convergence(self, r_proj, z_cl, z_src):
            \kappa = \frac{\Sigma}{\Sigma_{crit}}

        or
-
+        


m-aguena · 2021-11-04T16:15:14Z

clmm/theory/parent_class.py

@@ -374,3 +376,34 @@ def _eval_magnification(self, r_proj, z_cl, z_src):
        kappa = self.eval_convergence(r_proj, z_cl, z_src)
        gamma_t = self.eval_tangential_shear(r_proj, z_cl, z_src)
        return 1./((1-kappa)**2-abs(gamma_t)**2)
+
+    def eval_magnification_bias(self, r_proj, z_cl, z_src, alpha):


can you add an validation function? just follow the model of eval_magnification

m-aguena · 2021-11-04T16:16:04Z

clmm/theory/parent_class.py

@@ -2,7 +2,8 @@
 CLMModeling abstract class
 """
 import numpy as np
-from .generic import compute_reduced_shear_from_convergence
+import warnings
+from .generic import *


please do not import *

marina-ricci · 2021-11-12T15:08:46Z

@m-aguena , I think it should be good now!

…MM into issue/437/magnification_bias

m-aguena

@marina-ricci it looks good for me now.
I would just like @combet to have a final look before merging as I made a couple of changes.

marina-ricci · 2021-11-19T15:47:28Z

@hsinfan1996 Please be careful because the checks are failing since your commit.

hsinfan1996 · 2021-11-19T15:49:36Z

@hsinfan1996 Please be careful because the checks are failing since your commit.

I noticed, but I really don't think the check failures were related to my commit.

combet · 2021-11-19T15:55:56Z

The error appears when installing CLMM during CI (that's a new one...)

error: numpy 1.17.4 is installed but numpy>=1.18 is required by {'astropy'}
Error: Process completed with exit code 1.

hsinfan1996 · 2021-11-19T15:56:56Z

@combet @m-aguena astropy 5.0 is installed during install prereq using pip, since we only require astropy to be >=4. Too new.

hsinfan1996 · 2021-11-19T16:00:25Z

Need to edit requirements.txt.

combet

Thank you, all looks good!

combet and others added 5 commits June 11, 2021 17:01

Adding Einasto and Hernquist profiles from CCL. Note: the FFTLog inte…

b94f88b

…gration may still be optimized

Merge branch 'master' of https://github.com/LSSTDESC/CLMM

44af350

resolve merge conflict

afc5865

include magnification bias in generic.py

617506b

marina-ricci added the PR: Not Ready label Oct 1, 2021

marina-ricci added 15 commits October 25, 2021 12:23

add magnification bias in func layer

bcd73cf

Merge branch 'main' of https://github.com/LSSTDESC/CLMM into issue/43…

8544b4b

…7/magnification_bias

modify func call

1673025

modify func call

4a19d8b

added test for source in front of cluster

6af2f42

added oo function

538e03e

added test and call generic

90d3b43

corrected typo

61d2f04

corrected typo

dbd32a9

added import

2a52fe9

changed import

70e7585

corrected typo

a87ab01

corrected typo

7a16043

add mag bias in example thoery notebook

3c4c2ec

add mag bias in example thoery notebook oo

cb76e4a

typo?

7742c92

combet linked an issue Oct 26, 2021 that may be closed by this pull request

Include magnification bias in theory #437

Closed

3 tasks

marina-ricci added 2 commits October 27, 2021 11:00

commented test

b0880cd

commented test 2

4f3a056

marina-ricci added the PR: Ready for Review label Oct 28, 2021

m-aguena linked an issue Oct 28, 2021 that may be closed by this pull request

Check tolerance in theory test #453

Closed

combet reviewed Oct 29, 2021

View reviewed changes

marina-ricci added 2 commits October 29, 2021 16:02

add warning that does not seem to work

0aa8bdc

added warning

53d8c59

m-aguena requested changes Nov 4, 2021

View reviewed changes

marina-ricci added 3 commits November 12, 2021 15:47

reduce line more 100 char

4e7bd6c

remove extra tab and * import

5dc00c2

add validate function

132a2d7

marina-ricci and others added 3 commits November 12, 2021 17:06

resolve issue

0c155d8

use Modeling.eval_magnification_bias in compute_magnification_bias

c366395

remove h from Mpc/h

6957780

hsinfan1996 linked an issue Nov 19, 2021 that may be closed by this pull request

Mpc/h mentioned in demo_generate_mock_cluster notebook #458

Closed

m-aguena added 2 commits November 19, 2021 16:34

update tests

c6e53c5

Merge branch 'issue/437/magnification_bias' of github.com:LSSTDESC/CL…

bc2e4ce

…MM into issue/437/magnification_bias

m-aguena approved these changes Nov 19, 2021

View reviewed changes

require astropy<5

d26c340

combet approved these changes Nov 19, 2021

View reviewed changes

update version number

db63826

combet merged commit 41757b0 into main Nov 19, 2021

hsinfan1996 deleted the issue/437/magnification_bias branch March 17, 2022 15:42

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Issue/437/magnification bias #438

Issue/437/magnification bias #438

marina-ricci commented Oct 1, 2021

coveralls commented Oct 1, 2021 •

edited

Loading

marina-ricci commented Oct 25, 2021

marina-ricci commented Oct 25, 2021

m-aguena commented Oct 27, 2021

combet Oct 29, 2021 •

edited

Loading

marina-ricci Oct 29, 2021

hsinfan1996 Oct 29, 2021

combet Oct 29, 2021

hsinfan1996 Oct 29, 2021

marina-ricci Oct 29, 2021 •

edited

Loading

marina-ricci Nov 4, 2021 •

edited

Loading

m-aguena left a comment

m-aguena Nov 4, 2021

m-aguena Nov 4, 2021

m-aguena Nov 4, 2021

m-aguena Nov 4, 2021

m-aguena Nov 4, 2021

m-aguena Nov 4, 2021

marina-ricci commented Nov 12, 2021

m-aguena left a comment

marina-ricci commented Nov 19, 2021

hsinfan1996 commented Nov 19, 2021

combet commented Nov 19, 2021

hsinfan1996 commented Nov 19, 2021 •

edited

Loading

hsinfan1996 commented Nov 19, 2021

combet left a comment

		magnification_bias_from_magnification = np.array(
		magnification)**(np.array([alpha]).T - 1)

		@@ -281,7 +283,7 @@ def eval_convergence(self, r_proj, z_cl, z_src):
		\kappa = \frac{\Sigma}{\Sigma_{crit}}

		or

Issue/437/magnification bias #438

Issue/437/magnification bias #438

Conversation

marina-ricci commented Oct 1, 2021

coveralls commented Oct 1, 2021 • edited Loading

marina-ricci commented Oct 25, 2021

marina-ricci commented Oct 25, 2021

m-aguena commented Oct 27, 2021

combet Oct 29, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

marina-ricci Oct 29, 2021 • edited Loading

Choose a reason for hiding this comment

marina-ricci Nov 4, 2021 • edited Loading

Choose a reason for hiding this comment

m-aguena left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

marina-ricci commented Nov 12, 2021

m-aguena left a comment

Choose a reason for hiding this comment

marina-ricci commented Nov 19, 2021

hsinfan1996 commented Nov 19, 2021

combet commented Nov 19, 2021

hsinfan1996 commented Nov 19, 2021 • edited Loading

hsinfan1996 commented Nov 19, 2021

combet left a comment

Choose a reason for hiding this comment

coveralls commented Oct 1, 2021 •

edited

Loading

combet Oct 29, 2021 •

edited

Loading

marina-ricci Oct 29, 2021 •

edited

Loading

marina-ricci Nov 4, 2021 •

edited

Loading

hsinfan1996 commented Nov 19, 2021 •

edited

Loading