Source: https://github.com/aipoch/medical-research-skills
Survival Analysis (Kaplan-Meier)
Kaplan-Meier survival analysis tool for clinical and biological research. Generates publication-ready survival curves with statistical tests.
When to Use
- Use this skill when the task needs Kaplan-Meier survival analysis tool for clinical and biological research. Generates publication-ready survival curves with statistical tests.
- Use this skill for data analysis tasks that require explicit assumptions, bounded scope, and a reproducible output format.
- Use this skill when you need a documented fallback path for missing inputs, execution errors, or partial evidence.
Key Features
See ## Features above for related details.
- Scope-focused workflow aligned to: Kaplan-Meier survival analysis tool for clinical and biological research. Generates publication-ready survival curves with statistical tests.
- Packaged executable path(s):
scripts/main.py.
- Reference material available in
references/ for task-specific guidance.
- Structured execution path designed to keep outputs consistent and reviewable.
Dependencies
lifelines: Core survival analysis library
matplotlib, seaborn: Visualization
pandas, numpy: Data handling
scipy: Statistical tests
Example Usage
See ## Usage above for related details.
cd "20260318/scientific-skills/Data Analytics/survival-analysis-km"
python -m py_compile scripts/main.py
python scripts/main.py --help
Example run plan:
- Confirm the user input, output path, and any required config values.
- Edit the in-file
CONFIG block or documented parameters if the script uses fixed settings.
- Run
python scripts/main.py with the validated inputs.
- Review the generated output and return the final artifact with any assumptions called out.
Implementation Details
See ## Workflow above for related details.
- Execution model: validate the request, choose the packaged workflow, and produce a bounded deliverable.
- Input controls: confirm the source files, scope limits, output format, and acceptance criteria before running any script.
- Primary implementation surface:
scripts/main.py.
- Reference guidance:
references/ contains supporting rules, prompts, or checklists.
- Parameters to clarify first: input path, output path, scope filters, thresholds, and any domain-specific constraints.
- Output discipline: keep results reproducible, identify assumptions explicitly, and avoid undocumented side effects.
Quick Check
Use this command to verify that the packaged script entry point can be parsed before deeper execution.
python -m py_compile scripts/main.py
Audit-Ready Commands
Use these concrete commands for validation. They are intentionally self-contained and avoid placeholder paths.
python -m py_compile scripts/main.py
# Example invocation: python scripts/main.py --help
# Example invocation: python scripts/main.py --input "Audit validation sample with explicit symptoms, history, assessment, and next-step plan."
Workflow
- Confirm the user objective, required inputs, and non-negotiable constraints before doing detailed work.
- Validate that the request matches the documented scope and stop early if the task would require unsupported assumptions.
- Use the packaged script path or the documented reasoning path with only the inputs that are actually available.
- Return a structured result that separates assumptions, deliverables, risks, and unresolved items.
- If execution fails or inputs are incomplete, switch to the fallback path and state exactly what blocked full completion.
Features
- Kaplan-Meier Curve Generation: Publication-quality survival plots with confidence intervals
- Statistical Tests: Log-rank test, Wilcoxon test, Peto-Peto test
- Hazard Ratios: Cox proportional hazards regression with 95% CI
- Summary Statistics: Median survival time, restricted mean survival time (RMST)
- Multi-group Analysis: Supports 2+ comparison groups
- Risk Tables: Optional at-risk table below curves
Usage
Python Script
# Example invocation: python scripts/main.py --input data.csv --time time_col --event event_col --group group_col --output results/
Arguments
| Argument |
Description |
Required |
--input |
Input CSV file path |
Yes |
--time |
Column name for survival time |
Yes |
--event |
Column name for event indicator (1=event, 0=censored) |
Yes |
--group |
Column name for grouping variable |
Optional |
--output |
Output directory for results |
Yes |
--conf-level |
Confidence level (default: 0.95) |
Optional |
--risk-table |
Include risk table in plot |
Optional |
Input Format
CSV with columns:
- Time column: Numeric, time to event or censoring
- Event column: Binary (1 = event occurred, 0 = censored/right-censored)
- Group column: Categorical variable for stratification
Example:
patient_id,time_months,death,treatment_group
P001,24.5,1,Drug_A
P002,36.2,0,Drug_A
P003,18.7,1,Placebo
Output Files
km_curve.png: Kaplan-Meier survival curve
km_curve.pdf: Vector version for publications
survival_stats.csv: Statistical summary (median survival, confidence intervals)
hazard_ratios.csv: Cox regression results with HR and 95% CI
- `logrank_test.csv**: Pairwise comparison p-values
- `report.txt**: Human-readable summary report
Technical Details
Statistical Methods
Kaplan-Meier Estimator: Non-parametric maximum likelihood estimate of survival function
- Product-limit estimator: Ŝ(t) = Π(tᵢ≤t) (1 - dᵢ/nᵢ)
- Greenwood's formula for variance estimation
Log-Rank Test: Most widely used test for comparing survival curves
- Null hypothesis: No difference between groups
- Weighted by number at risk at each event time
Cox Proportional Hazards: Semi-parametric regression model
- h(t|X) = h₀(t) × exp(β₁X₁ + β₂X₂ + ...)
- Proportional hazards assumption checked via Schoenfeld residuals
Technical Difficulty: High ⚠️
This skill involves advanced statistical modeling. Results should be reviewed by a biostatistician, especially for:
- Proportional hazards assumption violations
- Small sample sizes (< 30 per group)
- Heavy censoring (> 50%)
- Time-varying covariates
References
See references/ folder for:
- Kaplan EL, Meier P (1958) original paper
- Cox DR (1972) regression models paper
- Sample datasets for testing
- Clinical reporting guidelines (ATN, CONSORT)
Parameters
| Parameter |
Type |
Default |
Description |
--input |
str |
Required |
Input CSV file path |
--time |
str |
Required |
Column name for survival time |
--event |
str |
Required |
|
--group |
str |
Required |
|
--output |
str |
Required |
Output directory for results |
--conf-level |
float |
0.95 |
|
--risk-table |
str |
Required |
Include risk table in plot |
--figsize |
str |
'10 |
|
--dpi |
int |
300 |
|
Example
# Basic survival curve
# Example invocation: python scripts/main.py \
--input clinical_data.csv \
--time overall_survival_months \
--event death \
--group treatment_arm \
--output ./results/ \
--risk-table
Output includes:
- Survival curves with 95% confidence bands
- Median survival: Drug A = 28.4 months (95% CI: 24.1-32.7), Placebo = 18.2 months (95% CI: 15.3-21.1)
- Log-rank test p-value: 0.0023
- Hazard ratio: 0.62 (95% CI: 0.45-0.85), p = 0.003
Risk Assessment
| Risk Indicator |
Assessment |
Level |
| Code Execution |
Python/R scripts executed locally |
Medium |
| Network Access |
No external API calls |
Low |
| File System Access |
Read input files, write output files |
Medium |
| Instruction Tampering |
Standard prompt guidelines |
Low |
| Data Exposure |
Output files saved to workspace |
Low |
Security Checklist
Prerequisites
# Python dependencies
pip install -r requirements.txt
Evaluation Criteria
Success Metrics
Test Cases
- Basic Functionality: Standard input → Expected output
- Edge Case: Invalid input → Graceful error handling
- Performance: Large dataset → Acceptable processing time
Lifecycle Status
- Current Stage: Draft
- Next Review Date: 2026-03-06
- Known Issues: None
- Planned Improvements:
- Performance optimization
- Additional feature support
Output Requirements
Every final response should make these items explicit when they are relevant:
- Objective or requested deliverable
- Inputs used and assumptions introduced
- Workflow or decision path
- Core result, recommendation, or artifact
- Constraints, risks, caveats, or validation needs
- Unresolved items and next-step checks
Error Handling
- If required inputs are missing, state exactly which fields are missing and request only the minimum additional information.
- If the task goes outside the documented scope, stop instead of guessing or silently widening the assignment.
- If
scripts/main.py fails, report the failure point, summarize what still can be completed safely, and provide a manual fallback.
- Do not fabricate files, citations, data, search results, or execution outcomes.
Input Validation
This skill accepts requests that match the documented purpose of survival-analysis-km and include enough context to complete the workflow safely.
Do not continue the workflow when the request is out of scope, missing a critical input, or would require unsupported assumptions. Instead respond:
survival-analysis-km only handles its documented workflow. Please provide the missing required inputs or switch to a more suitable skill.
Response Template
Use the following fixed structure for non-trivial requests:
- Objective
- Inputs Received
- Assumptions
- Workflow
- Deliverable
- Risks and Limits
- Next Checks
If the request is simple, you may compress the structure, but still keep assumptions and limits explicit when they affect correctness.
Inputs to Collect
- Required inputs: the user goal, the primary data or source file, and the requested output format.
- Optional inputs: output directory, formatting preferences, and validation constraints.
- If a required input is unavailable, return a short clarification request before continuing.
Output Contract
- Return a short summary, the main deliverables, and any assumptions that materially affect interpretation.
- If execution is partial, label what succeeded, what failed, and the next safe recovery step.
- Keep the final answer within the documented scope of the skill.
Validation and Safety Rules
- Validate identifiers, file paths, and user-provided parameters before execution.
- Do not fabricate results, metrics, citations, or downstream conclusions.
- Use safe fallback behavior when dependencies, credentials, or required inputs are missing.
- Surface any execution failure with a concise diagnosis and recovery path.
1---
2name: survival-analysis-km
3description: Kaplan-Meier survival analysis tool for clinical and biological research. Generates publication-ready survival curves with statistical tests.
4license: MIT
5author: AIPOCH
6metadata:
7 source: https://skillmd.com/skills/gabrielmoreira/survival-analysis-km
8---
9
10> **Source**: [https://github.com/aipoch/medical-research-skills](https://github.com/aipoch/medical-research-skills)
11
12# Survival Analysis (Kaplan-Meier)
13
14Kaplan-Meier survival analysis tool for clinical and biological research. Generates publication-ready survival curves with statistical tests.
15
16## When to Use
17
18- Use this skill when the task needs Kaplan-Meier survival analysis tool for clinical and biological research. Generates publication-ready survival curves with statistical tests.
19- Use this skill for data analysis tasks that require explicit assumptions, bounded scope, and a reproducible output format.
20- Use this skill when you need a documented fallback path for missing inputs, execution errors, or partial evidence.
21
22## Key Features
23
24See `## Features` above for related details.
25
26- Scope-focused workflow aligned to: Kaplan-Meier survival analysis tool for clinical and biological research. Generates publication-ready survival curves with statistical tests.
27- Packaged executable path(s): `scripts/main.py`.
28- Reference material available in `references/` for task-specific guidance.
29- Structured execution path designed to keep outputs consistent and reviewable.
30
31## Dependencies
32
33- `lifelines`: Core survival analysis library
34- `matplotlib`, `seaborn`: Visualization
35- `pandas`, `numpy`: Data handling
36- `scipy`: Statistical tests
37
38## Example Usage
39
40See `## Usage` above for related details.
41
42```bash
43cd "20260318/scientific-skills/Data Analytics/survival-analysis-km"
44python -m py_compile scripts/main.py
45python scripts/main.py --help
46```
47
48Example run plan:
491. Confirm the user input, output path, and any required config values.
502. Edit the in-file `CONFIG` block or documented parameters if the script uses fixed settings.
513. Run `python scripts/main.py` with the validated inputs.
524. Review the generated output and return the final artifact with any assumptions called out.
53
54## Implementation Details
55
56See `## Workflow` above for related details.
57
58- Execution model: validate the request, choose the packaged workflow, and produce a bounded deliverable.
59- Input controls: confirm the source files, scope limits, output format, and acceptance criteria before running any script.
60- Primary implementation surface: `scripts/main.py`.
61- Reference guidance: `references/` contains supporting rules, prompts, or checklists.
62- Parameters to clarify first: input path, output path, scope filters, thresholds, and any domain-specific constraints.
63- Output discipline: keep results reproducible, identify assumptions explicitly, and avoid undocumented side effects.
64
65## Quick Check
66
67Use this command to verify that the packaged script entry point can be parsed before deeper execution.
68
69```bash
70python -m py_compile scripts/main.py
71```
72
73## Audit-Ready Commands
74
75Use these concrete commands for validation. They are intentionally self-contained and avoid placeholder paths.
76
77```bash
78python -m py_compile scripts/main.py
79
80# Example invocation: python scripts/main.py --help
81
82# Example invocation: python scripts/main.py --input "Audit validation sample with explicit symptoms, history, assessment, and next-step plan."
83```
84
85## Workflow
86
871. Confirm the user objective, required inputs, and non-negotiable constraints before doing detailed work.
882. Validate that the request matches the documented scope and stop early if the task would require unsupported assumptions.
893. Use the packaged script path or the documented reasoning path with only the inputs that are actually available.
904. Return a structured result that separates assumptions, deliverables, risks, and unresolved items.
915. If execution fails or inputs are incomplete, switch to the fallback path and state exactly what blocked full completion.
92
93## Features
94
95- **Kaplan-Meier Curve Generation**: Publication-quality survival plots with confidence intervals
96- **Statistical Tests**: Log-rank test, Wilcoxon test, Peto-Peto test
97- **Hazard Ratios**: Cox proportional hazards regression with 95% CI
98- **Summary Statistics**: Median survival time, restricted mean survival time (RMST)
99- **Multi-group Analysis**: Supports 2+ comparison groups
100- **Risk Tables**: Optional at-risk table below curves
101
102## Usage
103
104### Python Script
105
106```text
107
108# Example invocation: python scripts/main.py --input data.csv --time time_col --event event_col --group group_col --output results/
109```
110
111### Arguments
112
113| Argument | Description | Required |
114|----------|-------------|----------|
115| `--input` | Input CSV file path | Yes |
116| `--time` | Column name for survival time | Yes |
117| `--event` | Column name for event indicator (1=event, 0=censored) | Yes |
118| `--group` | Column name for grouping variable | Optional |
119| `--output` | Output directory for results | Yes |
120| `--conf-level` | Confidence level (default: 0.95) | Optional |
121| `--risk-table` | Include risk table in plot | Optional |
122
123### Input Format
124
125CSV with columns:
126- **Time column**: Numeric, time to event or censoring
127- **Event column**: Binary (1 = event occurred, 0 = censored/right-censored)
128- **Group column**: Categorical variable for stratification
129
130Example:
131```csv
132patient_id,time_months,death,treatment_group
133P001,24.5,1,Drug_A
134P002,36.2,0,Drug_A
135P003,18.7,1,Placebo
136```
137
138### Output Files
139
140- `km_curve.png`: Kaplan-Meier survival curve
141- `km_curve.pdf`: Vector version for publications
142- `survival_stats.csv`: Statistical summary (median survival, confidence intervals)
143- `hazard_ratios.csv`: Cox regression results with HR and 95% CI
144- `logrank_test.csv**: Pairwise comparison p-values
145- `report.txt**: Human-readable summary report
146
147## Technical Details
148
149### Statistical Methods
150
1511. **Kaplan-Meier Estimator**: Non-parametric maximum likelihood estimate of survival function
152 - Product-limit estimator: Ŝ(t) = Π(tᵢ≤t) (1 - dᵢ/nᵢ)
153 - Greenwood's formula for variance estimation
154
1552. **Log-Rank Test**: Most widely used test for comparing survival curves
156 - Null hypothesis: No difference between groups
157 - Weighted by number at risk at each event time
158
1593. **Cox Proportional Hazards**: Semi-parametric regression model
160 - h(t|X) = h₀(t) × exp(β₁X₁ + β₂X₂ + ...)
161 - Proportional hazards assumption checked via Schoenfeld residuals
162
163### Technical Difficulty: High ⚠️
164
165This skill involves advanced statistical modeling. Results should be reviewed by a biostatistician, especially for:
166- Proportional hazards assumption violations
167- Small sample sizes (< 30 per group)
168- Heavy censoring (> 50%)
169- Time-varying covariates
170
171## References
172
173See `references/` folder for:
174- Kaplan EL, Meier P (1958) original paper
175- Cox DR (1972) regression models paper
176- Sample datasets for testing
177- Clinical reporting guidelines (ATN, CONSORT)
178
179## Parameters
180
181| Parameter | Type | Default | Description |
182|-----------|------|---------|-------------|
183| `--input` | str | Required | Input CSV file path |
184| `--time` | str | Required | Column name for survival time |
185| `--event` | str | Required | |
186| `--group` | str | Required | |
187| `--output` | str | Required | Output directory for results |
188| `--conf-level` | float | 0.95 | |
189| `--risk-table` | str | Required | Include risk table in plot |
190| `--figsize` | str | '10 | |
191| `--dpi` | int | 300 | |
192
193## Example
194
195```text
196
197# Basic survival curve
198
199# Example invocation: python scripts/main.py \
200 --input clinical_data.csv \
201 --time overall_survival_months \
202 --event death \
203 --group treatment_arm \
204 --output ./results/ \
205 --risk-table
206```
207
208Output includes:
209- Survival curves with 95% confidence bands
210- Median survival: Drug A = 28.4 months (95% CI: 24.1-32.7), Placebo = 18.2 months (95% CI: 15.3-21.1)
211- Log-rank test p-value: 0.0023
212- Hazard ratio: 0.62 (95% CI: 0.45-0.85), p = 0.003
213
214## Risk Assessment
215
216| Risk Indicator | Assessment | Level |
217|----------------|------------|-------|
218| Code Execution | Python/R scripts executed locally | Medium |
219| Network Access | No external API calls | Low |
220| File System Access | Read input files, write output files | Medium |
221| Instruction Tampering | Standard prompt guidelines | Low |
222| Data Exposure | Output files saved to workspace | Low |
223
224## Security Checklist
225
226- [ ] No hardcoded credentials or API keys
227- [ ] No unauthorized file system access (../)
228- [ ] Output does not expose sensitive information
229- [ ] Prompt injection protections in place
230- [ ] Input file paths validated (no ../ traversal)
231- [ ] Output directory restricted to workspace
232- [ ] Script execution in sandboxed environment
233- [ ] Error messages sanitized (no stack traces exposed)
234- [ ] Dependencies audited
235
236## Prerequisites
237
238```text
239
240# Python dependencies
241pip install -r requirements.txt
242```
243
244## Evaluation Criteria
245
246### Success Metrics
247- [ ] Successfully executes main functionality
248- [ ] Output meets quality standards
249- [ ] Handles edge cases gracefully
250- [ ] Performance is acceptable
251
252### Test Cases
2531. **Basic Functionality**: Standard input → Expected output
2542. **Edge Case**: Invalid input → Graceful error handling
2553. **Performance**: Large dataset → Acceptable processing time
256
257## Lifecycle Status
258
259- **Current Stage**: Draft
260- **Next Review Date**: 2026-03-06
261- **Known Issues**: None
262- **Planned Improvements**:
263 - Performance optimization
264 - Additional feature support
265
266## Output Requirements
267
268Every final response should make these items explicit when they are relevant:
269
270- Objective or requested deliverable
271- Inputs used and assumptions introduced
272- Workflow or decision path
273- Core result, recommendation, or artifact
274- Constraints, risks, caveats, or validation needs
275- Unresolved items and next-step checks
276
277## Error Handling
278
279- If required inputs are missing, state exactly which fields are missing and request only the minimum additional information.
280- If the task goes outside the documented scope, stop instead of guessing or silently widening the assignment.
281- If `scripts/main.py` fails, report the failure point, summarize what still can be completed safely, and provide a manual fallback.
282- Do not fabricate files, citations, data, search results, or execution outcomes.
283
284## Input Validation
285
286This skill accepts requests that match the documented purpose of `survival-analysis-km` and include enough context to complete the workflow safely.
287
288Do not continue the workflow when the request is out of scope, missing a critical input, or would require unsupported assumptions. Instead respond:
289
290> `survival-analysis-km` only handles its documented workflow. Please provide the missing required inputs or switch to a more suitable skill.
291
292## Response Template
293
294Use the following fixed structure for non-trivial requests:
295
2961. Objective
2972. Inputs Received
2983. Assumptions
2994. Workflow
3005. Deliverable
3016. Risks and Limits
3027. Next Checks
303
304If the request is simple, you may compress the structure, but still keep assumptions and limits explicit when they affect correctness.
305
306## Inputs to Collect
307
308- Required inputs: the user goal, the primary data or source file, and the requested output format.
309- Optional inputs: output directory, formatting preferences, and validation constraints.
310- If a required input is unavailable, return a short clarification request before continuing.
311
312## Output Contract
313
314- Return a short summary, the main deliverables, and any assumptions that materially affect interpretation.
315- If execution is partial, label what succeeded, what failed, and the next safe recovery step.
316- Keep the final answer within the documented scope of the skill.
317
318## Validation and Safety Rules
319
320- Validate identifiers, file paths, and user-provided parameters before execution.
321- Do not fabricate results, metrics, citations, or downstream conclusions.
322- Use safe fallback behavior when dependencies, credentials, or required inputs are missing.
323- Surface any execution failure with a concise diagnosis and recovery path.