Survival Analysis in R

Explore non-parametric survival models, including the Kaplan–Meier estimate and log-rank test, using survival object generated with serve function to relate time and event, comparing therapy groups in prostate cancer dataset.

Explore the survival function and its ties to hazard, cumulative hazard, and the Kaplan-Meier estimate, then compare non-parametric and parametric survival models with time-based graphs.

Master the survival object using the Surv function in R to encode time, event, and censoring (binary 0/1) for Kaplan-Meier plots and Cox models.

Explore the Kaplan-Meier estimator, a nonparametric method to estimate the survival function and its step plot, while accounting for censoring and comparing survival curves across groups.

Use the Kaplan-Meier method with the survival package to estimate survival from a Surv object. Plot a Kaplan-Meier curve with censoring marks and 95% confidence intervals.

Explore how to generate Kaplan-Meier survival plots in R using ggfortify and the CHIKI package, compare two treatment groups, adjust visuals, and interpret confidence intervals.

The log rank test, a nonparametric method, compares survival distributions between two groups using observed minus expected events at each event time and relates to Kaplan-Meier plots in clinical trials.

Explore the logrank test in R using the survival package, comparing two treatment groups, building a survival object, specifying time, status, and treatment, and interpreting p-values.

Apply the Kaplan-Meier estimator and log-rank test in R on the s.M Practicals dataset to assess temperature effects on survival, including a KM plot and a binary 180-degree variable.

Apply the Kaplan–Meier estimator and log-rank test in R's survival package to analyze time-to-failure data with censoring and a temperature covariate, revealing significant differences in survival between temperature groups.

Explore the Cox proportional hazards model in survival analysis, incorporating covariates to refine hazard rate estimates through the baseline hazard and an exponential covariate term.

Fit a Cox proportional hazards model in R with the survival package, using a survival object on the left and covariates on the right via a tilde formula, data=prostate.

Interpret the Cox proportional hazards results from the summary output, examine the concordance statistic and coefficients to assess covariate influence on survival, and view the plot around 65 months.

Learn how Aalen's additive regression model captures time dependent covariate effects, modeling cumulative hazard as alpha(t) plus X beta(t) and revealing evolving hazard patterns.

Explore the differences between parametric, nonparametric, and semi-parametric survival models, focusing on the survival object, baseline hazard, and covariates.

Learn to fit parametric survival models with flexible distributions using the Flex Serve package, comparing Wible and exponential fits and contrasting them with Kaplan-Meier plots.

Apply a Cox proportional hazards model to the Eureka UDCA dataset, interpret results, visualize survival between drug and placebo, and assess variable importance and concordance.

Learn to build and interpret a Cox proportional hazards model in R with the survival package, evaluating covariates such as TRT, stage, bilirubin, risk score, and concordance.

Explore survival trees, a decision-tree approach to modeling survival data, compare them with Cox models, and learn pruning to prevent overfitting on large datasets.

Explore how ranger constructs survival trees from survival data in R by using a formula with time, status, and covariates, and tailoring splits with various split rules.

Fit a survival tree using ranger with time and status as the input and covariates on the right, employing extra trees and permutation importance to derive an averaged survival curve.

Visualize the survival model in R by plotting the average survival probability over time as a step plot, and compare to Cox models while recognizing the small sample.

Compare the Kaplin Maior, Cox proportional hazards, and survival tree models using a unified comparison plot by extracting event times and survival probabilities into a single data frame.

Learn to build clean time columns for survival analysis by converting messy date-time data with base R (POSIXct/POSIXlt), the current package, and the lubricate package’s WRP time function.

Master dates and times in R, including time zones and formats, alongside leap years and seconds. Compare posixct, posixlt, and date, then use lubridate or chron for conversions.

Convert character date and time in a data frame to a date time column using ERP time, via a format string for uniform column format, yielding a POSIX t class.

learn how the lubridate package streamlines date and time handling in R for survival analysis, with ymd parsing, time zone management, and interval calculations.

Create a date time measurement data frame in R with date, time, and measurement columns, using seven values, converting formats to date and time, and rounding measurements to two decimals.

Master time calculations in lubridate by comparing minutes and duration versus period. Explain decimals vs integers, fuse units to seconds, and note leap year effects on date arithmetic.

Calculate days between entry date and last study visit in the UTSA 06 data using X1 (diff time) or X2 (interval with period conversion), and append as new time columns.

Explore missing value imputation and outlier detection to improve survival analysis, with theoretical underpinnings, practical methods, and implementations for small datasets.

Explain why data have missing values and distinguish missing completely at random, missing at random, and missing not at random, compare deletion, hot deck, mean imputation, interpolation, and multiple imputation.

Explore three levels of missing data handling in r: let functions handle na, delete rows with na.omit, and advanced imputation with packages like mice or input impute.

Learn how machine learning imputes missing data with the mice package in R, performing multiple imputation on a data frame, using random forest, and pooling results.

Identify how outliers affect results and handle them with three sigma rule, box plot rules, Q1 and Q3, and IQR, using model-based and proximity-based methods in survival analysis in R.

Detect outliers in univariate data using simple methods such as extreme studentized deviation intervals and box plot statistics, and confirm with Dixon or Grubbs tests from the outliers package.

Detect multivariate outliers in high-dimensional data with sign one, sign two, and pick out methods using principal component analysis distances; visualize results and interpret the W final indicator.

practice missing data imputation and outlier detection on the MGUS survival dataset, removing high-missing-value columns and imputing creatinine, albumin, and hemoglobin with mice.

Remove two index columns, identify creatinine outliers with a graps test, set outliers to NA, and impute missing data with mice using normal regression across three imputations for survival analysis.