[GSoC 2015 WIP] Testing other montecarlo functions #330

kaushik94 · 2015-06-12T11:13:18Z

@wkerzendorf is this how it should be done ?

kaushik94 · 2015-06-15T17:59:51Z

@wkerzendorf how should we go about testing the other functions that don't return anything ?

wkerzendorf · 2015-06-15T18:00:36Z

can you give an example?

kaushik94 · 2015-06-15T18:02:40Z

montecarlo_line_scatter ?

wkerzendorf · 2015-06-15T18:11:04Z

@kaushik94 that is a function that does the scatter of the montecarlo particles. What will change is the rpacket. This is some physics - can you make a PR just testing the montecarlo_line_scatter. @aoifeboyle can you help with the physics here.

kaushik94 · 2015-06-15T18:20:30Z

@wkerzendorf oh, In that case I guess all the functions in #295 need to be tested the same way. The trivial functions are tested in this PR. Can this be merged now ? And regarding the rest of the functions, I might need help from @aoifeboyle . @aoifeboyle please let me know a good time for you tomorrow, we can finish all of them off tomorrow.

aoifeboyle · 2015-06-15T19:00:23Z

@wkerzendorf @kaushik94 I'm not very clear on the montecarlo stuff but I can definitely try! Give me some time to have a look through it here. So should I just devise suitable tests for all of the functions?

wkerzendorf · 2015-06-15T19:04:23Z

@aoifeboyle - no specifically have a look at montecarlo_line_scatter. @kaushik94 will open a new PR on this soon and then we can think about what physics makes sense to test.

aoifeboyle · 2015-06-15T19:55:42Z

@wkerzendorf Ok, having a look through it now!

kaushik94 · 2015-06-20T18:40:05Z

tardis/montecarlo/src/test_cmontecarlo.c

+}
+
+double
+test_compute_distance2electron(){


I would like to test these functions. What should I start testing with ? Can anyone help me with simulating a decent rpacket and a storage model ? @aoifeboyle or @wkerzendorf .

It's quite late here now, but I can help tomorrow @kaushik94 :)

aoifeboyle · 2015-06-21T17:08:34Z

@kaushik94, I think I'll start by creating a test for the montecarlo_line_scatter function as you were planning to. I'm not that clear on C so I think that would be a good exercise to help me understand it. And I'll create an rpacket and storage_model that makes sense in terms of the physics to test it with. Does that sound ok for starters?

kaushik94 · 2015-06-21T17:50:04Z

@aoifeboyle that'd be great.

kaushik94 · 2015-06-22T11:40:50Z

@aoifeboyle You can just write the specs of rpacket and storage model on a doc somewhere and share it with me, I'll take care of the coding part. What do you think ?

aoifeboyle · 2015-06-22T11:57:17Z

Hi @kaushik94, sure! I'll create a document on google docs.

aoifeboyle · 2015-06-22T13:53:30Z

@kaushik94 I've updated our google doc with explanations of all of the parameters you asked for, and a bit of code to let you extract lines data from the atomic datasets for running your tests. Let me know if there is anything that isn't exactly clear! I hope this helps. If you need any other parameters explained, just put them in the google doc and I'll fill in explanations.

aoifeboyle · 2015-06-23T12:00:09Z

@kaushik94 I added the new values you asked for to the google doc. Ping me next time you add new ones. It doesn't notify me automatically.

aoifeboyle · 2015-06-24T11:14:57Z

@kaushik94 I got an email saying you'd left a comment here about your tests but I can't see it here - did you remove it? I looked at the tests in test_cmontecarlo.py anyway, and the only one that failed for me was test_rpacket_doppler_factor, but I was able to fix it by changing the doppler_factor variable for comparison slightly. Were you referring to something else? Or did you work it out?

kaushik94 · 2015-06-24T11:32:16Z

@aoifeboyle Yeah I mistakenly added a file that caused that error, I fixed it an deleted the comment. Are you getting an error still ? I just committed. The problem was that I din't declare js and nubars for the storage model. Seems fine now. I am trying to test montecarlo_one_packet_loop, what do you think ? Should I test any other functions before that ?

aoifeboyle · 2015-06-24T11:48:17Z

Ah ok that's fine! And yeah I don't think it really matters which one you do next. Let me know if there are any more parameters you need me to define :)

kaushik94 · 2015-06-24T11:50:00Z

@aoifeboyle see the doc, just updated.

aoifeboyle · 2015-06-29T18:21:19Z

Hi Kaushik,

How are you getting along now?

Aoife

Sent from Windows Mail

From: Kaushik Varanasimailto:[email protected]
Sent: ?Wednesday?, ?24? ?June? ?2015 ?12?:?50
To: tardis-sn/tardismailto:[email protected]
Cc: Aoife Boylemailto:[email protected]

@aoifeboylehttps://github.com/aoifeboyle see the doc, just updated.

Reply to this email directly or view it on GitHubhttps://github.com//pull/330#issuecomment-114842363.

kaushik94 · 2015-06-29T18:41:02Z

Hi @aoifeboyle I think I need your help, come on the doc if you are free now.

aoifeboyle · 2015-06-29T18:47:15Z

I can't. I have to give my sister my laptop for the next hour or so. But I'll have a look at the doc then.

Sent from Windows Mail

From: Kaushik Varanasimailto:[email protected]
Sent: ?Monday?, ?29? ?June? ?2015 ?19?:?41
To: tardis-sn/tardismailto:[email protected]
Cc: Aoife Boylemailto:[email protected]

Hi @aoifeboylehttps://github.com/aoifeboyle I think I need your help, come on the doc if you are free now.

Reply to this email directly or view it on GitHubhttps://github.com//pull/330#issuecomment-116792711.

aoifeboyle · 2015-06-29T20:50:19Z

@kaushik94 Done!

kaushik94 · 2015-06-30T00:00:43Z

@aoifeboyle strange because it seems to me that the problem is in thomson_scatter but I don't know why, it was running fine before. It most probably could be due to the montecarlo_one_packet function call but we can take a look at this later. For now I need some actual values.

kaushik94 · 2015-06-30T00:04:04Z

tardis/montecarlo/src/test_cmontecarlo.c

+	/* The following values need to be 
+	*  reassigned properly. By default
+	*  I assigned all 0.
+	*/


@aoifeboyle I know this might take a while so sorry. I need some sensible values for the following. You can reply in the doc or even comment them in-line and please explain the terms if you have a little time. I thought of looking up in the atomic data but couldn't find much about the storage models and etc, or if you are busy just let me know where I can get the data elsewhere. Thanks a lot.

@kaushik94 Having a look now. Let's try to keep the google doc more tidy (just for storing the parameter details) so it can be a reference for you. Let's keep the conversation on GitHub/email.

@kaushik94 They're in the doc now. Most of it wasn't in the atomic data, but I've given you reasonable values (I think). Let me know if you have any more problems.

aoifeboyle · 2015-06-30T08:26:35Z

@kaushik94 Did you see what I wrote about test in the doc? Your test works fine for me. So I don't know why it doesn't work for you? How do you actually run the tests?

kaushik94 · 2015-06-30T13:40:17Z

@aoifeboyle yes I saw, it was my mistake. The tests were running sometimes and sometimes fail, because I forgot to declare a function in C file which was giving weird errors. Now it all works fine. Thanks, will take a look at the doc.

kaushik94 · 2015-07-01T09:59:41Z

@aoifeboyle thanks, just see if I did everything correctly.

kaushik94 · 2015-07-06T12:11:47Z

@wkerzendorf review and merge.

wkerzendorf · 2015-07-06T13:30:59Z

@kaushik94 it doesn't pass. can you elaborate what doesn't work.

kaushik94 · 2015-07-06T13:33:26Z

It works fine on my local. see this https://travis-ci.org/tardis-sn/tardis/jobs/69728700. strange error, says current_packet_id is not a member of storage model.

kaushik94 · 2015-07-06T14:01:33Z

I think there is a bug on the travis side or I am doing something terribly wrong. I am trying to debug but please let me know if you understood @wkerzendorf .

wkerzendorf · 2015-07-06T14:34:11Z

@kaushik94 hmmm - try to clean your working directory with git clean -dfx and see if it still compiles with python setup.py test. as travis runs ubuntu that should be the same.

kaushik94 · 2015-07-06T14:38:29Z

Yes it does.

On Mon, Jul 6, 2015 at 8:04 PM, Wolfgang Kerzendorf <
[email protected]> wrote:

@kaushik94 https://github.com/kaushik94 hmmm - try to clean your
working directory with git clean -dfx and see if it still compiles with python
setup.py test. as travis runs ubuntu that should be the same.

—
Reply to this email directly or view it on GitHub
#330 (comment).

kaushik94 · 2015-07-06T14:40:01Z

@wkerzendorf I got the error. Check the storage.h file of the master in github. Someone removed current_packet_id parameter from the master.

wkerzendorf · 2015-07-06T14:41:24Z

@kaushik94 yes that makes sense have a look at #338 that did that to parallelize the code. Can you rebase onto the current master?

kaushik94 · 2015-07-06T14:42:55Z

Okay got it.

kaushik94 · 2015-07-07T08:08:29Z

@wkerzendorf I am still getting an error

wkerzendorf · 2015-07-07T08:10:55Z

that's right - it's now called packet_id have a look at the code.

kaushik94 · 2015-07-07T10:55:50Z

@wkerzendorf please review and merge.

updating master

kaushik94 · 2015-07-16T10:58:29Z

Please don't merge, more functions to be added.

kaushik94 · 2015-07-27T18:30:24Z

@wkerzendorf please merge this. For reference #295 . I have not done testing for macro_atom because I needed more data.

kaushik94 · 2015-07-27T19:42:10Z

@wkerzendorf I am getting different answers for bf_cross_section on my local and travis. Travis expects 0 while on my local its -1. Can you take a look into the test_bf_cross_section function, I marked it with xfail.

kaushik94 · 2015-07-27T23:00:31Z

@wkerzendorf please merge this. But I don't know how to go about from here. A little help is much required. Thanks.

[GSoC 2015 WIP] Testing other montecarlo functions

kaushik94 mentioned this pull request Jun 15, 2015

GSoC 2015 - Testing #295

Closed

20 tasks

kaushik94 mentioned this pull request Jun 17, 2015

[GSoC 2015 WIP] test_montecarlo_line_scatter #334

Closed

kaushik94 reviewed Jun 20, 2015
View reviewed changes

kaushik94 reviewed Jun 30, 2015
View reviewed changes

kaushik94 added 2 commits July 16, 2015 16:05

Testing all functions in cmontecarlo

5ad3c64

updating master

adding continuum function

a703bc2

kaushik94 added 5 commits July 27, 2015 23:23

testing bound_free and chi_bf

98324af

removing unnecessary

037425a

removing binary

47f4cb1

test bf_cross_section

5588b58

free free scatter

96ea0ac

kaushik94 added 2 commits July 28, 2015 00:57

removing bi_cross_section

ece4548

marking xfails

8d1c9fd

wkerzendorf added a commit that referenced this pull request Aug 3, 2015

Merge pull request #330 from kaushik94/other-montecarlo

48dc3aa

[GSoC 2015 WIP] Testing other montecarlo functions

wkerzendorf merged commit 48dc3aa into tardis-sn:master Aug 3, 2015

[GSoC 2015 WIP] Testing other montecarlo functions #330

[GSoC 2015 WIP] Testing other montecarlo functions #330

Conversation

kaushik94 commented Jun 12, 2015

kaushik94 commented Jun 15, 2015

wkerzendorf commented Jun 15, 2015

kaushik94 commented Jun 15, 2015

wkerzendorf commented Jun 15, 2015

kaushik94 commented Jun 15, 2015

aoifeboyle commented Jun 15, 2015

wkerzendorf commented Jun 15, 2015

aoifeboyle commented Jun 15, 2015

kaushik94 Jun 20, 2015

Choose a reason for hiding this comment

aoifeboyle Jun 20, 2015

Choose a reason for hiding this comment

aoifeboyle commented Jun 21, 2015

kaushik94 commented Jun 21, 2015

kaushik94 commented Jun 22, 2015

aoifeboyle commented Jun 22, 2015

aoifeboyle commented Jun 22, 2015

aoifeboyle commented Jun 23, 2015

aoifeboyle commented Jun 24, 2015

kaushik94 commented Jun 24, 2015

aoifeboyle commented Jun 24, 2015

kaushik94 commented Jun 24, 2015

aoifeboyle commented Jun 29, 2015

kaushik94 commented Jun 29, 2015

aoifeboyle commented Jun 29, 2015

aoifeboyle commented Jun 29, 2015

kaushik94 commented Jun 30, 2015

kaushik94 Jun 30, 2015

Choose a reason for hiding this comment

aoifeboyle Jun 30, 2015

Choose a reason for hiding this comment

aoifeboyle Jun 30, 2015

Choose a reason for hiding this comment

aoifeboyle commented Jun 30, 2015

kaushik94 commented Jun 30, 2015

kaushik94 commented Jul 1, 2015

kaushik94 commented Jul 6, 2015

wkerzendorf commented Jul 6, 2015

kaushik94 commented Jul 6, 2015

kaushik94 commented Jul 6, 2015

wkerzendorf commented Jul 6, 2015

kaushik94 commented Jul 6, 2015

kaushik94 commented Jul 6, 2015

wkerzendorf commented Jul 6, 2015

kaushik94 commented Jul 6, 2015

kaushik94 commented Jul 7, 2015

wkerzendorf commented Jul 7, 2015

kaushik94 commented Jul 7, 2015

kaushik94 commented Jul 16, 2015

kaushik94 commented Jul 27, 2015

kaushik94 commented Jul 27, 2015

kaushik94 commented Jul 27, 2015