Course Update: Using Cursor for Coding Demonstration

1. The Dynamic Prediction Model Course Based on R Language Will Use Cursor for Coding and Course Instruction

2. The Course Will Demonstrate and Assist Students in Completing Cursor Configuration

Why Use Cursor

AI Models Can Be Invoked Anytime

Rich AI Models (ctrl+L)

Unlimited Low-Price Memberships Can Be Purchased Only from Certain Platforms

Automatically Generate Code Based on Comments

Generate Code According to Code Block Requirements (ctrl+K)

SSH Login to Remote Server R Language Environment

Workspace Similar to R Studio

Supports Shortcut Key Settings Similar to R Studio

Advantages and Disadvantages of Using R with Cursor

• Advantages:
• Extremely Powerful Interface for AI Coding Workflow
• Very Good Git GUI Tool
• Powerful Cross-Language Tools for JSON Parsing, GitHub Operations, and Dockerfile Scripts
• Disadvantages:
• Annoying to Use with renv

Summary

1. Low-Cost Use of AI-Assisted Coding Tools
2. Seamless Transition with R Studio Experience, AI Assistance at 10x Speed
3. Multi-threading Based on VSCode
4. Can Also Serve as a Python Editor
5. Rich Plugins, Strong Scalability

Course Objectives and Introduction

As an important part of real-world research, predictive models are widely studied. However, traditional predictive models use baseline data to predict the final survival outcomes, which cannot incorporate important data that may dynamically change during subsequent follow-ups (such as dynamic changes in tumor markers). This situation can lead to estimation bias in statistics and is inconsistent with clinical reality. Recently developed dynamic prediction model methods utilize multiple follow-up data of patients, combined with baseline data, to estimate the final survival outcomes of patients (or similar time-to-event events). The volume of published articles is rapidly increasing.

In clinical practice, doctors make further diagnostic and treatment judgments based on patients’ dynamic indicators. The dynamic prediction model combines longitudinal data of patients with final survival outcomes for more accurate predictions of the final results. Since R language currently occupies an important position in medical statistical work, many clinical doctors and nurses find it difficult to combine R language with clinical research due to time constraints. Therefore, the R Language Dynamic Prediction Model Course is designed to quickly enable students to master the commonly used R language in statistical work, assisting clinical research. Tianqi Zhuli (Tianjin) Productivity Promotion Co., Ltd. is hosting the “R Language-Based Dynamic Prediction Model Course Training Class”.

The Development of Predictive Model Articles Can Be Summarized into Three Stages:

1. Traditional epidemiological nomogram models (essentially Cox regression and GLM regression), simple statistical analysis models, are model-dependent methods. It is difficult to meet the premise assumptions in clinical practice, resulting in poor actual effects.

2. Construction of predictive models based on machine learning/deep learning (increasing dimensions in data and introducing machine learning in algorithms) has made data processing more flexible and reduced model assumptions. However, the data used still relies on a single baseline data for patients, which is inconsistent with clinical reality.

3. Dynamic prediction models based on longitudinal data (using longitudinal multiple follow-up data, applying joint models and other dynamic prediction model methods) predict final survival outcomes based on multiple follow-up data of patients, aligning more closely with clinical reality in both data and methodology.

Considering the characteristics of dynamic prediction models, they are essential methods for subsequent high-scoring articles:

1. Data must include multiple follow-up data for the same patient, which is more challenging to collect compared to past cross-sectional baseline data. Additionally, dynamic prediction models require fitting longitudinal linear mixed models, thus necessitating a larger data volume. This suggests that if such data can be collected, it will be easier to publish high-scoring articles.

2. To apply the methodology of dynamic prediction models, one must first master ordinary survival analysis and general predictive model methods, and also be familiar with generalized linear mixed models for longitudinal data analysis. Furthermore, one needs to understand the tidyverse syntax basics to transform their data into longitudinal data that meets function requirements. Additionally, for joint models, the forms of model combination and variable selection must be considered from both clinical background and statistical methodology perspectives.

Recent Examples of High-Scoring Articles

Instructor

Flexible Fatty – Alone

PhD in Oncology from a Double First-Class University, currently employed at one of the top five cancer centers in China. Research focuses on real-world research, bioinformatics analysis, and artificial intelligence research. Currently published over 10 SCI papers as the first or co-first author, with a cumulative IF of 50+. Currently collaborating on research with multiple domestic universities and hospitals. Jointly translated the full text of jmbayes2 and dynamicLM into Chinese for the first time in the country and published it on the public account.

Course Directory and Schedule

Teaching Format and Time

Teaching Format: Remote Online Real-Time Live Teaching.

Teaching Time: Starting in December 2024, with a total of no less than 30 hours, 3-5 hours of teaching each week, allowing ample time for learning, expected to complete all teaching content in 6-8 weeks.

Q&A Support: A dedicated WeChat group for the course will be established for free Q&A support for one year.

Video Replay: Free unlimited replays within 3 years.

Course Price and After-Sales Guarantee

Course Price: Total 3000 yuan, registration can be made with an advance payment of 300 yuan, and the remaining amount can be paid within 2 weeks after the course starts

Please contact the teaching assistant in advance for public transfer and other procedures.

Organizing Company: Tianqi Zhuli (Tianjin) Productivity Promotion Co., Ltd.

Incentive Policy: Tuition can be refunded if students publish articles with IF 10+ using the learned content (specific requirements and processes need to be consulted with the teaching assistant).

Registration Inquiry

You Can Contact My Teaching Assistant for Consultation

Teaching Assistant Contact Number: 18502623993

Official Notification

PDF version of the notification can be obtained by contacting the teaching assistant