Manimaran Pandiyan’s Post

Analysis dataset Programming with R. ADSL - then OCCD dataset, now time for BDS : Exposure. #adam #admiral #R #BDS #ADEX https://lnkd.in/gXYD8NNQ

Hi Data- Scientists Today I programmed in Spyder a new validation method for factor analysis. Steps are: 1. Use a factor- analysis with has just one output: a labda - matrix L. I programmed a new Principal Axis Factor analysis (PAF) with just that, and only that, output . 2. calculate the generalised Inverse (g_i) of scaled datamatrix DM and store it as transpose i_DM = g_i(DM).T 3. take a random 1/ 3 sample of DM , with replace = False, and calculate L - matrices of IN - sample en OUT - sample parts of DM 4. calculate i_DM * L , with * as matrixmultiplication, for both L - matrices 5. calculate the correlations between the two matrices of step 4. The important theoretical step is the possibility I discovered of calcalating factorscores Fs by Fs = i_DM * L happy coding!

We hope to see you at SESUG 2024 in Bethesda, Maryland! Join us at 8 a.m., September 24, for “Efficient Ways to Create Adverse Event Summary Tables by System Organ Class and Preferred Term using PROC SQL in SAS,” a presentation by Hardik Sheth, Manager, Statistical Programming, Catalyst Flex. Discover Catalyst Flex’s comprehensive solutions and experienced teams of biostatisticians and programmers. And learn how we can help support your clinical research needs. #Biometrics #SAS #ConnectWithCatalyst

🚨 Blink and you will miss it! 🚨 Join Phastar’s Daniel Lindenbaum at PHUSE EU Connect as he unlocks the potential of SAS hash table programming! Discover how to leverage this powerful tool for faster, more efficient data processing and simplifying tasks 🔎 📍 Wednesday 13th November, 12:00 – 12:30pm CET #Phastar #PHUSE2024 #SASProgramming #DataEfficiency #DataScience

Which R function is the closest to data frame options? For filtering records and selecting variables, R programmers can apply the subset function or data frame options to get the same results! #opensource #R #Rlearning #opendata #cdisc #sdtm #adam #submissions #dataquality #RGuru #RTraining #Rclass #SAS2R #SASSavvy #PharmaSUG #WUSS #PHUSE #Pharmaverse #SunilGupta #RcenterofExcellence

I completed the R Fundamental for Data Science course, and it was a valuable experience that gave me a strong foundation in R programming. The course started with an introduction to R, where I learned the basics of setting up RStudio and writing simple scripts. Then, I delved into working with vectors, lists, and data frames, essential data manipulation structures. One of the highlights was exploring charts and graphics in R, where I learned how to create various visualizations to represent data effectively. The hands-on practice helped solidify my understanding of these concepts.

Sometimes we use imblearn to solve imbalance in classification tasks. But do we know how to use it during cross-validation? If you are resampling before cross validation, this will cause data leakage. It changes the class distribution in the data you will test and you will get unrealistic validation scores. You can create a separate sampler object for each fold. Although this method works, your code will not be very readable. Moreover, you cannot embed this method in the cross_val_score function in sklearn. You need to write your own for-loop. The method I recommend is as follows. You can build a pipeline with sampler and classifier. Pipeline ensures that sampling is applied only to the training data during cross-validation.

You can access the exercise at the R for Fun Club on afredac.net. Also don’t miss this beginner class by joining here https://lnkd.in/e2F_gAXV Knowledge of R programming will help you in data cleaning, analysis and visualization. The previous sessions are available at https://lnkd.in/e5qWiaDg #afredac #rforfunclub #rfc #rprogrammer #rprogramming #data #datascience