HitOrMissMonteCarlo Test Added with Minor change in timeout setting #35
Conversation
|
@tdhock @faizan-khan-iit I've added a test for Hit or Miss Monte Carlo Integration and Pi value estimation. |
| x1_nodes <- getNodeSet(info$html, "//g[@class='geom2_point_vizone']/g[@class='PANEL1']/circle/@cx") | ||
| y1_nodes <- getNodeSet(info$html, "//g[@class='geom2_point_vizone']/g[@class='PANEL1']/circle/@cy") | ||
|
|
||
| Sys.sleep(1.739) |
There was a problem hiding this comment.
I took the reference from other successful tests. Sys.sleep() is used to pause the program for the specified time so that next reading can be taken. I used the same time as in other tests.
|
great start but would be even better if you could edit the .travis.yml config to fix the build system error. I reported the issue here ropensci/RSelenium#230 but no response. maybe you could try upgrading the version of RSelenium used in DESCRIPTION? |
|
Okay, sure thing! |
|
|
||
| coordPanel2 <- getNodeSet(info$newHtml, "//g[@class='geom5_path_viztwo']/g[@class='PANEL1']/path/@d") | ||
|
|
||
| expect_false(identical(coordPanel1, coordPanel2)) |
There was a problem hiding this comment.
Does this test assume that getNodeSet will always return the rendered points in the same order? I would prefer not to check it this way because even if the plot is not rendering any new points if getNodeSet returns them in a different order, this test might pass. Though if you are sure about the ordering, no changes needed. To me, the length of the retrieved nodes is a better indicator.
There was a problem hiding this comment.
I guess checking the length of retrieved nodes is a better indicator. I'll make changes and push as soon as possible.
There was a problem hiding this comment.
Hii @tdhock, @faizan-khan-iit,
I observed that the points are being rendered in order.
I checked and found this about the order.
So, the identical check is good to go. But if you say I can move to length check which will also work!
|
typically i the order does not matter then you can sort the
…On Sun, Mar 29, 2020 at 8:47 AM Himanshu Singh ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In tests/testthat/test-renderer3-HitOrMissMonteCarlo.R
<#35 (comment)>:
> + y2_nodes <- getNodeSet(info$newHtml, ***@***.******@***.******@***.***")
+
+ expect_false(identical(x1_nodes, x2_nodes))
+ expect_false(identical(y1_nodes, y2_nodes))
+})
+
+test_that("Different points are rendered on Pi estimation plot", {
+ coordPanel1 <- getNodeSet(info$html, ***@***.******@***.******@***.***")
+
+ Sys.sleep(1.739)
+
+ info$newHtml <- getHTML()
+
+ coordPanel2 <- getNodeSet(info$newHtml, ***@***.******@***.******@***.***")
+
+ expect_false(identical(coordPanel1, coordPanel2))
I guess checking the length of retrieved nodes is a better indicator. I'll
make changes and push as soon as possible.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#35 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAHDX4VXZK7KCIHQVD4O2SLRJ5UQPANCNFSM4LGUQCEQ>
.
|
@tdhock, Looks like the error didn't occur this time which occurred last time. But different error occurred this time which is in renderer5 and compiler. Have a look please? https://travis-ci.org/github/tdhock/animint2/builds/674748257 |
I've added HitOrMissMonteCarlo Test for Hard Test of
Animated interactive ggplotsI had to change the setImplicitTimeout method as it wasn't working and seems to be defunct.