Laust Mortensen

Sign up for Post Prod! post.news/?r=eVfmI

Having interesting discussions on #interpretability and different models at @DSTdk with @udansk! Thanks @needs_project for making it possible! #riseneeds #XAI

What joy to log on to twitter and not find brain numbing amounts of COVID-related crap, just brain numbing amounts of old school twitter crap!

@Doc_Courtney Sorry -- I did not know that back story and just wanted to say that I think that you are doing very well if those are your issues in your work life. I suspect that you are a really good mentor.

@udansk Between my job and raising a child by myself (my former husband died when he was five), I have zero time for self care and it’s a first world problem?

Professor friends, any suggestions for striking a balance between editing/commenting on mentees’ work versus working on your own writing? Maybe I just have too many productive mentees, but I’m drowning.

FYI: simple benchmark (330k rows, 65k groups, 6 cols) using latest CRAN binaries of {collapse}, {dplyr} & {data.table}. Results: (1) {collapse} is also competitive to {data.table} with base R functions. (2) {dplyr} code can run <= 100x faster using {collapse}. #RStats #tidyverse

Yes, COVID is on par with Flu in terms of risk to kids, but more to the point, the amount of restrictions we placed on children is in no way whatsoever commensurate with their risk of COVID. We restructured their entire lives, and deprived them of a year of education. Horrible

@LyngseF #FearTheBanana !

The work was completed while I was in Italy getting a nice SARS-CoV-2 infections while skiing. Poetic justice after leaving my coauthors with the work, I guess😎

@LyngseF But...

@udansk Nemesis!

THIS IS THE KEY POINT: BA.2 > BA.1 > DELTA. I AM SO UPSET I CAN'T STOP YELLING.

@StreetCred2017 @LyngseF @CarstenKirkeby @KareMolbak @TyraGroveKrause @RskSkov @mobdjek @MarcStegger @Economics_UCPH @CEBI_UCPH @ssi We did not, but boy did we wish that we had the data to do just that. Colleagues @SSI_dk are hard at work to do this, but it will take a little while.

@LyngseF @udansk @CarstenKirkeby @KareMolbak @TyraGroveKrause @RskSkov @mobdjek @MarcStegger @Economics_UCPH @CEBI_UCPH @ssi Thank you! Since they are co-circulating, Any opinion on protection offered by BA.1 infections against BA.2?

New Preprint Out! “Transmission of SARS-CoV-2 Omicron VOC subvariants BA.1 and BA.2: Evidence from Danish Households” We investigate the secondary attack rate (SAR) in #Omicron BA.1 vs. BA.2 households in Demark, 20 December 2021 – 18 January 2022. medrxiv.org/content/10.110… 1/

@JosetteSchoenma @LyngseF @CarstenKirkeby @KareMolbak @TyraGroveKrause @RskSkov @mobdjek @MarcStegger @Economics_UCPH @CEBI_UCPH @ssi The point of the exercise is to show how the results might differ if analyzed differently, i.e. the robustness of the analysis. We definitely like the analyses beneath Table 3 the best.

@JosetteSchoenma @LyngseF @CarstenKirkeby @KareMolbak @TyraGroveKrause @RskSkov @mobdjek @MarcStegger @Economics_UCPH @CEBI_UCPH @ssi Table 14 is based on models that are run on stratified subsamples of the whole dataset: Stratification by the vaccination status of the primary and potential secondary cases. This means that the estimates comes with slightly fewer assumptions, but will be excepted to vary more.

To set the stage, here is the 14-day SAR for different Variant of Concerns (VOC). “New” #Omicron BA.2 (red): 42% “Original” #Omicron BA.1 (blue): 35% #Delta VOC (purple): 33% Unknown VOC (no WGS results on sample) (gray): 33% 3/

@Sunshin47139354 @C_Siemering @LyngseF @SpiessKatja @AndersFomsgaard @MortenRas71 @MarcStegger @OvervadMaria @KareMolbak @TyraGroveKrause Eeek! But if we - contrary to fact - had found that the paper would have been real easy to publis, I guess🤔

@udansk @C_Siemering @LyngseF @SpiessKatja @AndersFomsgaard @MortenRas71 @MarcStegger @OvervadMaria @KareMolbak @TyraGroveKrause I think the question was asked because some people now say: "See, science has found that with Omicron, vaccinated people carry the double risk of infection compared to non-vaccinated people".

1/n Mal wieder ein schönes Beispiel für Manipulation seitens Homburg. Suggeriert werden soll mal wieder, dass Impfungen schädlich sind. Die Studie sagt jedoch das Gegenteil.

@C_Siemering @LyngseF @SpiessKatja @AndersFomsgaard @MortenRas71 @MarcStegger @OvervadMaria @KareMolbak @TyraGroveKrause OK, got it. Perhaps we - the authors - should make the study more abuse-proof by e.g. reporting in absolute probability of catching SARS-CoV2 in the six groups compared? The odds ratios are perhaps too easy to misinterpret.

@udansk @LyngseF @SpiessKatja @AndersFomsgaard @MortenRas71 @MarcStegger @OvervadMaria @KareMolbak @TyraGroveKrause This is also my take-away. Homburg is known for spreading misleading information about effectiveness of vaccines and is talking much about deaths by corona-vaccines. His quote deliberately distorts the results of the study as it suggests that SAR increases due to vaccination.

@C_Siemering @LyngseF @SpiessKatja @AndersFomsgaard @MortenRas71 @MarcStegger @OvervadMaria @KareMolbak @TyraGroveKrause Well, I would'n be able to understand the findings without context. But that is also true for the discussion between you and Homburg. Pre-print take-away for me: vaccines/boosting works well, just not as well for omicron due to immune evasiveness

10/n Let's ask the authors. How do you comment on Homburg's abbreviated quote from your study? @LyngseF, @udansk, @SpiessKatja, @AndersFomsgaard, @MortenRas71, @MarcStegger, @OvervadMaria, @KareMolbak, @TyraGroveKrause

🚨EURO #PhDSchool on #DataDriven #DecisionMaking and #Optimization, June 13-22, 2022, IMUS-Instituto de Matemáticas de la Universidad de Sevilla, @imus_es 📢Registration Open congreso.us.es/epsdata/ congreso.us.es/epsdata/docs/E… @EUROonline_News @CBS_ECON @needs_project @MSCActions