batch-cohort

Name: batch-cohort
Author: Aperivue

208stars53forks

How to install batch-cohort?

aperivue/medsci-skills/batch-cohort

$npx -y skills add Aperivue/medsci-skills --skill batch-cohort

Installs into the current project.

Summary

Generate N analysis scripts from a single methodology template × multiple exposure/outcome combinations. The "80-person team" pattern — same validated method, swap variables only. Produces batch R/Python code + summary matrix.

Files · 1

View on GitHub

SKILL.md

1	# Batch Cohort Analysis Skill
2
3	You are assisting a medical researcher in generating multiple analysis scripts from a single
4	validated methodology template, each differing only in the exposure/outcome variable combination.
5	This replicates the "80-person research team" pattern: one PI designs the methodology, and
6	many researchers execute the same approach with different variable swaps.
7
8	## When to Use
9
10	- Researcher has a validated analysis template (e.g., from /replicate-study or /cross-national)
11	- Wants to explore multiple exposure → outcome combinations on the same database
12	- Goal: systematic variable-swap code generation + batch execution + result matrix
13
14	## Inputs
15
16	1. Database path(s): CSV/SAS data files (KNHANES, NHANES, NHIS, or any cleaned cohort)
17	2. Methodology template: One of:
18	- Path to a validated R/Python analysis script (from /replicate-study or /cross-national)
19	- A paper type template name: `nhis_cohort`, `cross_national`, `survey_weighted`
20	- A source paper to extract methodology from (falls back to /replicate-study Phase 1)
21	3. Combination spec: A list of exposure/outcome pairs, provided as:
22	- Inline list: `exposures: [depression, obesity, smoking]; outcomes: [diabetes, hypertension, CVD]`
23	- CSV file with columns: `exposure`, `outcome`, (optional) `subgroup_vars`
24	- `"all"` keyword: generates all pairwise combinations from the lists
25
26	### Optional Inputs
27
28	- Covariate set: Fixed covariate list for all analyses (default: use template's set)
29	- Subgroup variables: Variables to stratify by (default: sex, age group)
30	- Output format: `code_only` (just scripts) \| `execute` (run + collect results) \| `full` (code + results + summary)
31	- Cross-national mode: If TRUE, generates paired scripts for both countries per combination
32
33	## Workflow
34
35	### Phase 1: Template Validation
36
37	1. Read the methodology template (R script or paper type reference).
38	2. Identify the slot variables — parts that change per combination:
39	- `EXPOSURE_VAR`: raw variable name in the database
40	- `EXPOSURE_LABEL`: human-readable label for tables/figures
41	- `EXPOSURE_CODING`: how to derive binary/categorical exposure
42	- `OUTCOME_VAR`: raw variable name
43	- `OUTCOME_LABEL`: human-readable label
44	- `OUTCOME_CODING`: how to derive binary outcome
45	3. Verify the template runs successfully on at least one combination before batch generation.
46	4. Output: template summary with identified slots → user approval.
47
48	### Phase 2: Variable Specification
49
50	For each exposure and outcome in the combination spec:
51
52	1. Look up the variable in the database:
53	- KNHANES: check variable name exists in the CSV header
54	- NHANES: check which table contains the variable (use codebook.csv if available)
55	- NHIS: check claims code or variable name
56	2. Define coding:
57	- Binary: threshold or category mapping (e.g., `HE_glu >= 126 → diabetes = 1`)
58	- Categorical: level definitions (e.g., `smoking: current/former/never`)
59	3. Check covariate overlap: If the exposure IS one of the standard covariates, remove it from the adjustment set for that analysis (no self-adjustment).
60	4. Output: combination matrix with all variable specifications.
61
62	```
63	\| # \| Exposure \| Exposure Coding \| Outcome \| Outcome Coding \| Covariates (adjusted) \| Notes \|
64	\|---\|----------\|-----------------\|---------\|----------------\|----------------------\|-------\|
65	\| 1 \| Depression (PHQ≥10) \| BP_PHQ sum ≥10 \| Diabetes \| HE_glu≥126\|HbA1c≥6.5\|DE1_dg=1 \| age,sex,edu,income,smoking,alcohol,obesity,CVD \| — \|
66	\| 2 \| Obesity (BMI≥25) \| HE_obe ≥4 \| Diabetes \| same \| age,sex,edu,income,smoking,alcohol,depression,CVD \| obesity removed from covariates \|
67	\| ... \| \| \| \| \| \| \|
68	```
69
70	### Phase 3: Batch Code Generation
71
72	For each combination in the matrix:
73
74	1. Clone the template script.
75	2. Replace slot variables with the combination-specific values.
76	3. Adjust covariates: Remove exposure variable from covariate list if present.
77	4. Set output paths: Each combination gets its own results subdirectory.
78	5. Generate a master runner script (`run_all.R` or `run_all.sh`) that:
79	- Executes all N scripts sequentially (or in parallel via `future`/`parallel`)
80	- Captures errors per script without stopping the batch
81	- Logs execution time per analysis
82
83	### Phase 4: Batch Execution (if `execute` or `full` mode)
84
85	1. Run the master script.
86	2. Collect results from each combination's output directory.
87	3. Handle failures gracefully:
88	- Log which combinations failed

People who liked this also viewed

6 picks

Type

batch-cohort

208stars53forks

View on GitHub

How to install batch-cohort?

aperivue/medsci-skills/batch-cohort

$npx -y skills add Aperivue/medsci-skills --skill batch-cohort

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Batch Cohort Analysis Skill
2
3	You are assisting a medical researcher in generating multiple analysis scripts from a single
4	validated methodology template, each differing only in the exposure/outcome variable combination.
5	This replicates the "80-person research team" pattern: one PI designs the methodology, and
6	many researchers execute the same approach with different variable swaps.
7
8	## When to Use
9
10	- Researcher has a validated analysis template (e.g., from /replicate-study or /cross-national)
11	- Wants to explore multiple exposure → outcome combinations on the same database
12	- Goal: systematic variable-swap code generation + batch execution + result matrix
13
14	## Inputs
15
16	1. Database path(s): CSV/SAS data files (KNHANES, NHANES, NHIS, or any cleaned cohort)
17	2. Methodology template: One of:
18	- Path to a validated R/Python analysis script (from /replicate-study or /cross-national)
19	- A paper type template name: `nhis_cohort`, `cross_national`, `survey_weighted`
20	- A source paper to extract methodology from (falls back to /replicate-study Phase 1)
21	3. Combination spec: A list of exposure/outcome pairs, provided as:
22	- Inline list: `exposures: [depression, obesity, smoking]; outcomes: [diabetes, hypertension, CVD]`
23	- CSV file with columns: `exposure`, `outcome`, (optional) `subgroup_vars`
24	- `"all"` keyword: generates all pairwise combinations from the lists
25
26	### Optional Inputs
27
28	- Covariate set: Fixed covariate list for all analyses (default: use template's set)
29	- Subgroup variables: Variables to stratify by (default: sex, age group)
30	- Output format: `code_only` (just scripts) \| `execute` (run + collect results) \| `full` (code + results + summary)
31	- Cross-national mode: If TRUE, generates paired scripts for both countries per combination
32
33	## Workflow
34
35	### Phase 1: Template Validation
36
37	1. Read the methodology template (R script or paper type reference).
38	2. Identify the slot variables — parts that change per combination:
39	- `EXPOSURE_VAR`: raw variable name in the database
40	- `EXPOSURE_LABEL`: human-readable label for tables/figures
41	- `EXPOSURE_CODING`: how to derive binary/categorical exposure
42	- `OUTCOME_VAR`: raw variable name
43	- `OUTCOME_LABEL`: human-readable label
44	- `OUTCOME_CODING`: how to derive binary outcome
45	3. Verify the template runs successfully on at least one combination before batch generation.
46	4. Output: template summary with identified slots → user approval.
47
48	### Phase 2: Variable Specification
49
50	For each exposure and outcome in the combination spec:
51
52	1. Look up the variable in the database:
53	- KNHANES: check variable name exists in the CSV header
54	- NHANES: check which table contains the variable (use codebook.csv if available)
55	- NHIS: check claims code or variable name
56	2. Define coding:
57	- Binary: threshold or category mapping (e.g., `HE_glu >= 126 → diabetes = 1`)
58	- Categorical: level definitions (e.g., `smoking: current/former/never`)
59	3. Check covariate overlap: If the exposure IS one of the standard covariates, remove it from the adjustment set for that analysis (no self-adjustment).
60	4. Output: combination matrix with all variable specifications.
61
62	```
63	\| # \| Exposure \| Exposure Coding \| Outcome \| Outcome Coding \| Covariates (adjusted) \| Notes \|
64	\|---\|----------\|-----------------\|---------\|----------------\|----------------------\|-------\|
65	\| 1 \| Depression (PHQ≥10) \| BP_PHQ sum ≥10 \| Diabetes \| HE_glu≥126\|HbA1c≥6.5\|DE1_dg=1 \| age,sex,edu,income,smoking,alcohol,obesity,CVD \| — \|
66	\| 2 \| Obesity (BMI≥25) \| HE_obe ≥4 \| Diabetes \| same \| age,sex,edu,income,smoking,alcohol,depression,CVD \| obesity removed from covariates \|
67	\| ... \| \| \| \| \| \| \|
68	```
69
70	### Phase 3: Batch Code Generation
71
72	For each combination in the matrix:
73
74	1. Clone the template script.
75	2. Replace slot variables with the combination-specific values.
76	3. Adjust covariates: Remove exposure variable from covariate list if present.
77	4. Set output paths: Each combination gets its own results subdirectory.
78	5. Generate a master runner script (`run_all.R` or `run_all.sh`) that:
79	- Executes all N scripts sequentially (or in parallel via `future`/`parallel`)
80	- Captures errors per script without stopping the batch
81	- Logs execution time per analysis
82
83	### Phase 4: Batch Execution (if `execute` or `full` mode)
84
85	1. Run the master script.
86	2. Collect results from each combination's output directory.
87	3. Handle failures gracefully:
88	- Log which combinations failed

People who liked this also viewed

6 picks

Type