Performance and trendy updates#172
Conversation
…ifier is provided
…ng full removal (deprecated) or support (experimental)
- delta2kernel - ffill_na in trend - R code for running trends, kernels, etc. All in C now.
- delta2kernel - ffill_na in trend - R code for running trends, kernels, etc. All in C now.
…requires update of tests
…into oo_refactor_simd
|
Arghh - something broke, sorry, I thought check passed but it didn't... |
…Allows for keeping track of which columns are design columns and which are parameter names. Also allows for more robust checking of duplicate parameter names.
|
Alright, fixed. @niekstevenson do you want to press the magic button? |
…he y-label said density. Fixed by plotting density by default. If prob=FALSE is passed, it relabels to 'count' correctly. Update tests accordingly.
|
Hey I can press the magic button, but it seems like you're still making commits? It's looking very clean as always! Just let me know when you're ready |
|
Yes - again, not sure what the best way of working here is :-) Reason I'm still making commits is that I am working on one of the tutorials, for which I need this code. The commits pertain things not really related to the big merge itself - e.g., minor plotting changes, etc. |
|
Sounds good! For now I will merge this into dev and you can branch of that again? |
|
yey!
… On 22 May 2026, at 14:09, niekstevenson ***@***.***> wrote:
niekstevenson
left a comment
(ampl-psych/EMC2#172)
<#172 (comment)>
Sounds good! For now I will merge this into dev and you can branch of that again?
—
Reply to this email directly, view it on GitHub <#172 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEHU7ESQ5RDXJK453JT25UD44A7P7AVCNFSM6AAAAACZGEPHKWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DKMJYGUYTKMBWGY>.
Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.
You are receiving this because you authored the thread.
|
StevenM1
force-pushed
the
oo_refactor_simd
branch
4 times, most recently
from
bef2af3 to
48924ff
Compare
…mfcol. By default setting par() with mfcol/mfrow overwrites the cex setting, which limits user control over label sizes
StevenM1
force-pushed
the
oo_refactor_simd
branch
from
48924ff to
8bd4038
Compare
StevenM1
force-pushed
the
oo_refactor_simd
branch
2 times, most recently
from
811e095 to
91a4fcb
Compare
… overlaying ground-truth (data-generating) trajectory/ies
StevenM1
force-pushed
the
oo_refactor_simd
branch
from
91a4fcb to
6288de4
Compare
2 participants
The big one. Includes:
A couple of important things to be aware of:
By default, it longer relies on R's pnorm, but instead uses Hart's approximation (~double precision). can be reverted back to original R's pnorm with a compile-time flag. See
pnorm_utils.h.For macOS, makes use of the Accelerate platform to speed up some transformations (exp and log). This requires a
configurationinstead of justMakevars(which don't allow for running arbitrary bash code, which is needed to detect the platform). configure is basically a .sh script that generates Makevars. The defaults should be CRAN-safe. For optimal performance, useremotes::install_github('ampl-psych/EMC2@dev, configure.args=c('--native'))