PICdb-MEDS 0.0.1__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- picdb_meds-0.0.1/LICENSE +21 -0
- picdb_meds-0.0.1/PKG-INFO +166 -0
- picdb_meds-0.0.1/README.md +150 -0
- picdb_meds-0.0.1/pyproject.toml +68 -0
- picdb_meds-0.0.1/setup.cfg +4 -0
- picdb_meds-0.0.1/src/PICdb_MEDS/__init__.py +12 -0
- picdb_meds-0.0.1/src/PICdb_MEDS.egg-info/PKG-INFO +166 -0
- picdb_meds-0.0.1/src/PICdb_MEDS.egg-info/SOURCES.txt +10 -0
- picdb_meds-0.0.1/src/PICdb_MEDS.egg-info/dependency_links.txt +1 -0
- picdb_meds-0.0.1/src/PICdb_MEDS.egg-info/entry_points.txt +2 -0
- picdb_meds-0.0.1/src/PICdb_MEDS.egg-info/requires.txt +1 -0
- picdb_meds-0.0.1/src/PICdb_MEDS.egg-info/top_level.txt +1 -0
picdb_meds-0.0.1/LICENSE
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2026 Robin van de Water
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
|
13
|
+
copies or substantial portions of the Software.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
21
|
+
SOFTWARE.
|
|
@@ -0,0 +1,166 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: PICdb-MEDS
|
|
3
|
+
Version: 0.0.1
|
|
4
|
+
Summary: An ETL pipeline to extract the Paediatric Intensive Care (PIC) database into the MEDS format.
|
|
5
|
+
Author-email: Robin van de Water <rpv2114@cumc.columbia.edu>
|
|
6
|
+
Project-URL: Homepage, https://github.com/rvandewater/PICdb_MEDS
|
|
7
|
+
Project-URL: Issues, https://github.com/rvandewater/PICdb_MEDS/issues
|
|
8
|
+
Classifier: Programming Language :: Python :: 3
|
|
9
|
+
Classifier: License :: OSI Approved :: MIT License
|
|
10
|
+
Classifier: Operating System :: OS Independent
|
|
11
|
+
Requires-Python: <3.14,>=3.11.4
|
|
12
|
+
Description-Content-Type: text/markdown
|
|
13
|
+
License-File: LICENSE
|
|
14
|
+
Requires-Dist: meds-extract[download]<0.8,>=0.7.0
|
|
15
|
+
Dynamic: license-file
|
|
16
|
+
|
|
17
|
+
# PICdb MEDS Extraction ETL
|
|
18
|
+
|
|
19
|
+
<!-- Placeholders until available: the PyPI badges resolve once PICdb-MEDS is published; replace
|
|
20
|
+
CODECOV_TOKEN_PLACEHOLDER with the repo's Codecov badge token and zenodo.XXXXXXX with its Zenodo DOI. -->
|
|
21
|
+
|
|
22
|
+
[](https://pypi.org/project/PICdb-MEDS/)
|
|
23
|
+
[](https://codecov.io/gh/rvandewater/PICdb_MEDS)
|
|
24
|
+
[](https://github.com/rvandewater/PICdb_MEDS/actions/workflows/synthetic-etl.yaml)
|
|
25
|
+
[](https://github.com/rvandewater/PICdb_MEDS/actions/workflows/code-quality-main.yaml)
|
|
26
|
+
|
|
27
|
+
|
|
28
|
+
[](https://github.com/rvandewater/PICdb_MEDS#license)
|
|
29
|
+
[](https://github.com/rvandewater/PICdb_MEDS/pulls)
|
|
30
|
+
[](https://github.com/rvandewater/PICdb_MEDS/graphs/contributors)
|
|
31
|
+
[](https://doi.org/10.5281/zenodo.XXXXXXX)
|
|
32
|
+
|
|
33
|
+
This pipeline extracts the [Paediatric Intensive Care (PIC) database v1.1.0](https://physionet.org/content/picdb/1.1.0/)
|
|
34
|
+
from PhysioNet into the [MEDS format](https://medical-event-data-standard.github.io/). PIC holds 12,881
|
|
35
|
+
neonatal and paediatric patients and 13,449 hospital admissions from the Children's Hospital of Zhejiang
|
|
36
|
+
University School of Medicine (2010–2018). It was built from the
|
|
37
|
+
[ETL_MEDS_Template](https://github.com/Medical-Event-Data-Standard/ETL_MEDS_Template) and uses the
|
|
38
|
+
config-only layout introduced by [MEDS-Extract](https://github.com/mmcdermott/MEDS_extract) 0.7.
|
|
39
|
+
|
|
40
|
+
## Usage
|
|
41
|
+
|
|
42
|
+
PIC is a credentialed dataset. You need a PhysioNet account that has signed the PIC data use agreement.
|
|
43
|
+
|
|
44
|
+
```bash
|
|
45
|
+
pip install git+https://github.com/rvandewater/PICdb_MEDS.git
|
|
46
|
+
export DATASET_DOWNLOAD_USERNAME=$PHYSIONET_USERNAME
|
|
47
|
+
export DATASET_DOWNLOAD_PASSWORD=$PHYSIONET_PASSWORD
|
|
48
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR
|
|
49
|
+
```
|
|
50
|
+
|
|
51
|
+
When you run this, the program will:
|
|
52
|
+
|
|
53
|
+
1. Download the v1.1.0 tables (`V1.1.0/*.csv`, about 1.2 GB) from PhysioNet into
|
|
54
|
+
`$MEDS_OUTPUT_DIR/.meds_extract_run/raw_input`. Files that already exist and match their checksums
|
|
55
|
+
are skipped, so an interrupted download resumes. Pass `download_dest_dir=$RAW_INPUT_DIR` to keep the
|
|
56
|
+
raw data somewhere durable and reuse it across runs.
|
|
57
|
+
2. Build the MEDS cohort from those files and write it to `$MEDS_OUTPUT_DIR` (`data/` and `metadata/`).
|
|
58
|
+
All transformations, joins and metadata extraction are declared in
|
|
59
|
+
[`src/PICdb_MEDS/event_configs.yaml`](src/PICdb_MEDS/event_configs.yaml).
|
|
60
|
+
|
|
61
|
+
If you already have the PhysioNet release on disk, skip the download and point `input_dir` at the
|
|
62
|
+
release root, the directory that contains `V1.1.0/`:
|
|
63
|
+
|
|
64
|
+
```bash
|
|
65
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR \
|
|
66
|
+
do_download=false input_dir=/path/to/paediatric-intensive-care-database-1.1.0
|
|
67
|
+
```
|
|
68
|
+
|
|
69
|
+
To parallelize the stages, pass a runner file:
|
|
70
|
+
|
|
71
|
+
```bash
|
|
72
|
+
printf 'parallelize:\n n_workers: 8\n launcher: joblib\n' >runner.yaml
|
|
73
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR stage_runner_fp=runner.yaml
|
|
74
|
+
```
|
|
75
|
+
|
|
76
|
+
Run `meds-extract-run --help` for every option.
|
|
77
|
+
|
|
78
|
+
## How this ETL is defined
|
|
79
|
+
|
|
80
|
+
The package contains no Python. The whole ETL is one MESSY (MEDS-Extract Specification Syntax YAML)
|
|
81
|
+
file, [`src/PICdb_MEDS/event_configs.yaml`](src/PICdb_MEDS/event_configs.yaml). It declares:
|
|
82
|
+
|
|
83
|
+
- **`sources:`**: where the raw data lives (the PhysioNet release, restricted to `V1.1.0/*.csv`).
|
|
84
|
+
- **`etl:`**: the dataset name and sharding options.
|
|
85
|
+
- **One block per raw table**: how each row becomes one or more MEDS events, and which dictionary
|
|
86
|
+
tables describe the resulting codes.
|
|
87
|
+
|
|
88
|
+
`pyproject.toml` registers the file with MEDS-Extract's `MEDS_extract.pipelines` entry-point group under
|
|
89
|
+
the name `PICdb`. That registration is what makes `spec=PICdb` work. No pre-MEDS step is needed: every
|
|
90
|
+
PIC table already carries `SUBJECT_ID`, and every timestamp has the same format.
|
|
91
|
+
|
|
92
|
+
## Events
|
|
93
|
+
|
|
94
|
+
| Table | Event codes | Notes |
|
|
95
|
+
| --------------------- | -------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------- |
|
|
96
|
+
| `PATIENTS` | `GENDER//{gender}`, `MEDS_BIRTH`, `MEDS_DEATH` | Death time is the earliest `ADMISSIONS.DEATHTIME`, falling back to `DOD`. |
|
|
97
|
+
| `ADMISSIONS` | `HOSPITAL_ADMISSION//{dept}`, `HOSPITAL_DISCHARGE//{dept}`, `INSURANCE//…`, `ETHNICITY//…`, `ADMISSION_DIAGNOSIS//ICD10_CN//{code}` | ED times equal admit/discharge times in every row, so there are no separate ED events. |
|
|
98
|
+
| `ICUSTAYS` | `ICU_ADMISSION//{careunit}`, `ICU_DISCHARGE//{careunit}` | |
|
|
99
|
+
| `DIAGNOSES_ICD` | `DIAGNOSIS//ICD10_CN//{code}` | Placed at hospital discharge. `seq_num` and `diag_category` are kept as columns. |
|
|
100
|
+
| `LABEVENTS` | `LAB//{itemid}//{unit}` | `numeric_value`, `text_value` and `flag` are kept. |
|
|
101
|
+
| `CHARTEVENTS` | `CHART//{itemid}//{unit}` | |
|
|
102
|
+
| `INPUTEVENTS` | `FLUID_INPUT//{unit}` | PIC records only fluid intake amounts here, with no item ID. |
|
|
103
|
+
| `OUTPUTEVENTS` | `OUTPUT//{itemid}//{unit}` | |
|
|
104
|
+
| `PRESCRIPTIONS` | `MEDICATION//START//{drug}//{dose unit}`, `MEDICATION//END//{drug}` | Uses the English drug name. The dose is the `numeric_value`. |
|
|
105
|
+
| `MICROBIOLOGYEVENTS` | `MICROBIOLOGY//SPECIMEN//{spec_itemid}`, `MICROBIOLOGY//ORGANISM//{org_itemid}`, `MICROBIOLOGY//SUSCEPTIBILITY//{ab_itemid}//{S/I/R…}` | The MIC dilution is the `numeric_value`. Organism and specimen IDs are kept as columns. |
|
|
106
|
+
| `OR_EXAM_REPORTS` | `EXAM//{type}//{item}` | Uses the exam time, or the report time when the exam time is missing. |
|
|
107
|
+
| `EMR_SYMPTOMS` | `SYMPTOM//{symptom}//{+/-}` | Symptoms extracted by NLP from clinical notes. |
|
|
108
|
+
| `SURGERY_INFO` | `SURGERY//START//{name}`, `SURGERY//END//{name}`, `ANESTHESIA//START//{method}`, `ANESTHESIA//END//{method}` | |
|
|
109
|
+
| `SURGERY_VITAL_SIGNS` | `SURGERY_VITAL//{itemid}` | Intra-operative monitoring. Units are in the code metadata (`unit`). |
|
|
110
|
+
|
|
111
|
+
A missing unit or susceptibility interpretation becomes `UNK` in the code. A row with a missing
|
|
112
|
+
identifying component (for example, no organism) produces no event for that component.
|
|
113
|
+
|
|
114
|
+
### Code metadata
|
|
115
|
+
|
|
116
|
+
`metadata/codes.parquet` links codes to PIC's dictionary tables:
|
|
117
|
+
|
|
118
|
+
- **`description`**: the English label from `D_ITEMS`, `D_LABITEMS` or `D_ICD_DIAGNOSES`.
|
|
119
|
+
- **`parent_codes`**: `LOINC/…` for lab items with a well-formed LOINC code, and `ICD10/…` for
|
|
120
|
+
diagnoses whose mapped code is a valid 3- or 4-character WHO ICD-10 code. PIC's own extended codes
|
|
121
|
+
(7+ characters) have no parent.
|
|
122
|
+
- **Chinese source terms**: kept in their own columns (`title_cn`, `label_cn`, `drug_name_cn`,
|
|
123
|
+
`org_name_cn`, `spec_type_desc_cn`, `ab_name_cn`, `symptom_name_cn`). Codes are always built from the
|
|
124
|
+
English or ID columns.
|
|
125
|
+
- **Extra dictionary fields**: `fluid`, `category` and `unit`.
|
|
126
|
+
|
|
127
|
+
### Data notes
|
|
128
|
+
|
|
129
|
+
- **Shifted dates.** Dates are shifted 50–100 years into the future per patient. Years between 2049 and
|
|
130
|
+
2119 are expected, and intervals within a patient are preserved.
|
|
131
|
+
- **Release layout.** The v1.1.0 release ships the current tables under `V1.1.0/`. The gzipped files at
|
|
132
|
+
the release root are the superseded v1.0.0 tables, and this ETL does not read them.
|
|
133
|
+
- **Swapped column names.** In `PRESCRIPTIONS`, the CSV header `DRUG_NAME` holds the Chinese name and
|
|
134
|
+
`DRUG_NAME_EN` the English one. The bundled SQL scripts name these columns differently.
|
|
135
|
+
- **Many-to-many microbiology IDs.** Item IDs and their Chinese names do not map one-to-one, so codes
|
|
136
|
+
use the item IDs. Expect MEDS-Extract warnings that some codes have several `*_cn` values; those
|
|
137
|
+
values are aggregated into lists.
|
|
138
|
+
- **Uninformative columns.** `LANGUAGE`, `RELIGION` and `MARITAL_STATUS` hold a single value across the
|
|
139
|
+
release and are not extracted.
|
|
140
|
+
- **Local diagnosis codes.** Diagnoses use PIC's extended Chinese ICD-10 codes (`ICD10_CN`), so they do
|
|
141
|
+
not yet line up with standard ICD-10 codes in other datasets. Valid standard codes are attached as
|
|
142
|
+
`parent_codes`. Mapping equivalent codes to standard ICD-10 needs a MEDS-Extract feature; see
|
|
143
|
+
[#1](https://github.com/rvandewater/PICdb_MEDS/issues/1).
|
|
144
|
+
|
|
145
|
+
## Testing
|
|
146
|
+
|
|
147
|
+
PIC has no open demo release. Instead, `tests/e2e_synthetic_test.py` writes a few fully synthetic rows
|
|
148
|
+
with PIC's exact headers and layout, runs `meds-extract-run spec=PICdb` over them, and checks the
|
|
149
|
+
events, times, values and code metadata it produces.
|
|
150
|
+
|
|
151
|
+
```bash
|
|
152
|
+
uv sync --group dev
|
|
153
|
+
uv run pytest tests/ -v
|
|
154
|
+
uv run pre-commit run --all-files
|
|
155
|
+
```
|
|
156
|
+
|
|
157
|
+
## Citation
|
|
158
|
+
|
|
159
|
+
If you use this ETL, please cite the PIC database and its descriptor paper:
|
|
160
|
+
|
|
161
|
+
- Li, H., Zeng, X., & Yu, G. (2020). *Paediatric Intensive Care database* (version 1.1.0). PhysioNet.
|
|
162
|
+
https://doi.org/10.13026/32x9-wv38
|
|
163
|
+
- Zeng, X., Yu, G., Lu, Y. et al. PIC, a paediatric-specific intensive care database. *Sci Data* 7, 14
|
|
164
|
+
(2020). https://doi.org/10.1038/s41597-020-0355-4
|
|
165
|
+
|
|
166
|
+
See also [`CITATION.cff`](CITATION.cff).
|
|
@@ -0,0 +1,150 @@
|
|
|
1
|
+
# PICdb MEDS Extraction ETL
|
|
2
|
+
|
|
3
|
+
<!-- Placeholders until available: the PyPI badges resolve once PICdb-MEDS is published; replace
|
|
4
|
+
CODECOV_TOKEN_PLACEHOLDER with the repo's Codecov badge token and zenodo.XXXXXXX with its Zenodo DOI. -->
|
|
5
|
+
|
|
6
|
+
[](https://pypi.org/project/PICdb-MEDS/)
|
|
7
|
+
[](https://codecov.io/gh/rvandewater/PICdb_MEDS)
|
|
8
|
+
[](https://github.com/rvandewater/PICdb_MEDS/actions/workflows/synthetic-etl.yaml)
|
|
9
|
+
[](https://github.com/rvandewater/PICdb_MEDS/actions/workflows/code-quality-main.yaml)
|
|
10
|
+
|
|
11
|
+
|
|
12
|
+
[](https://github.com/rvandewater/PICdb_MEDS#license)
|
|
13
|
+
[](https://github.com/rvandewater/PICdb_MEDS/pulls)
|
|
14
|
+
[](https://github.com/rvandewater/PICdb_MEDS/graphs/contributors)
|
|
15
|
+
[](https://doi.org/10.5281/zenodo.XXXXXXX)
|
|
16
|
+
|
|
17
|
+
This pipeline extracts the [Paediatric Intensive Care (PIC) database v1.1.0](https://physionet.org/content/picdb/1.1.0/)
|
|
18
|
+
from PhysioNet into the [MEDS format](https://medical-event-data-standard.github.io/). PIC holds 12,881
|
|
19
|
+
neonatal and paediatric patients and 13,449 hospital admissions from the Children's Hospital of Zhejiang
|
|
20
|
+
University School of Medicine (2010–2018). It was built from the
|
|
21
|
+
[ETL_MEDS_Template](https://github.com/Medical-Event-Data-Standard/ETL_MEDS_Template) and uses the
|
|
22
|
+
config-only layout introduced by [MEDS-Extract](https://github.com/mmcdermott/MEDS_extract) 0.7.
|
|
23
|
+
|
|
24
|
+
## Usage
|
|
25
|
+
|
|
26
|
+
PIC is a credentialed dataset. You need a PhysioNet account that has signed the PIC data use agreement.
|
|
27
|
+
|
|
28
|
+
```bash
|
|
29
|
+
pip install git+https://github.com/rvandewater/PICdb_MEDS.git
|
|
30
|
+
export DATASET_DOWNLOAD_USERNAME=$PHYSIONET_USERNAME
|
|
31
|
+
export DATASET_DOWNLOAD_PASSWORD=$PHYSIONET_PASSWORD
|
|
32
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR
|
|
33
|
+
```
|
|
34
|
+
|
|
35
|
+
When you run this, the program will:
|
|
36
|
+
|
|
37
|
+
1. Download the v1.1.0 tables (`V1.1.0/*.csv`, about 1.2 GB) from PhysioNet into
|
|
38
|
+
`$MEDS_OUTPUT_DIR/.meds_extract_run/raw_input`. Files that already exist and match their checksums
|
|
39
|
+
are skipped, so an interrupted download resumes. Pass `download_dest_dir=$RAW_INPUT_DIR` to keep the
|
|
40
|
+
raw data somewhere durable and reuse it across runs.
|
|
41
|
+
2. Build the MEDS cohort from those files and write it to `$MEDS_OUTPUT_DIR` (`data/` and `metadata/`).
|
|
42
|
+
All transformations, joins and metadata extraction are declared in
|
|
43
|
+
[`src/PICdb_MEDS/event_configs.yaml`](src/PICdb_MEDS/event_configs.yaml).
|
|
44
|
+
|
|
45
|
+
If you already have the PhysioNet release on disk, skip the download and point `input_dir` at the
|
|
46
|
+
release root, the directory that contains `V1.1.0/`:
|
|
47
|
+
|
|
48
|
+
```bash
|
|
49
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR \
|
|
50
|
+
do_download=false input_dir=/path/to/paediatric-intensive-care-database-1.1.0
|
|
51
|
+
```
|
|
52
|
+
|
|
53
|
+
To parallelize the stages, pass a runner file:
|
|
54
|
+
|
|
55
|
+
```bash
|
|
56
|
+
printf 'parallelize:\n n_workers: 8\n launcher: joblib\n' >runner.yaml
|
|
57
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR stage_runner_fp=runner.yaml
|
|
58
|
+
```
|
|
59
|
+
|
|
60
|
+
Run `meds-extract-run --help` for every option.
|
|
61
|
+
|
|
62
|
+
## How this ETL is defined
|
|
63
|
+
|
|
64
|
+
The package contains no Python. The whole ETL is one MESSY (MEDS-Extract Specification Syntax YAML)
|
|
65
|
+
file, [`src/PICdb_MEDS/event_configs.yaml`](src/PICdb_MEDS/event_configs.yaml). It declares:
|
|
66
|
+
|
|
67
|
+
- **`sources:`**: where the raw data lives (the PhysioNet release, restricted to `V1.1.0/*.csv`).
|
|
68
|
+
- **`etl:`**: the dataset name and sharding options.
|
|
69
|
+
- **One block per raw table**: how each row becomes one or more MEDS events, and which dictionary
|
|
70
|
+
tables describe the resulting codes.
|
|
71
|
+
|
|
72
|
+
`pyproject.toml` registers the file with MEDS-Extract's `MEDS_extract.pipelines` entry-point group under
|
|
73
|
+
the name `PICdb`. That registration is what makes `spec=PICdb` work. No pre-MEDS step is needed: every
|
|
74
|
+
PIC table already carries `SUBJECT_ID`, and every timestamp has the same format.
|
|
75
|
+
|
|
76
|
+
## Events
|
|
77
|
+
|
|
78
|
+
| Table | Event codes | Notes |
|
|
79
|
+
| --------------------- | -------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------- |
|
|
80
|
+
| `PATIENTS` | `GENDER//{gender}`, `MEDS_BIRTH`, `MEDS_DEATH` | Death time is the earliest `ADMISSIONS.DEATHTIME`, falling back to `DOD`. |
|
|
81
|
+
| `ADMISSIONS` | `HOSPITAL_ADMISSION//{dept}`, `HOSPITAL_DISCHARGE//{dept}`, `INSURANCE//…`, `ETHNICITY//…`, `ADMISSION_DIAGNOSIS//ICD10_CN//{code}` | ED times equal admit/discharge times in every row, so there are no separate ED events. |
|
|
82
|
+
| `ICUSTAYS` | `ICU_ADMISSION//{careunit}`, `ICU_DISCHARGE//{careunit}` | |
|
|
83
|
+
| `DIAGNOSES_ICD` | `DIAGNOSIS//ICD10_CN//{code}` | Placed at hospital discharge. `seq_num` and `diag_category` are kept as columns. |
|
|
84
|
+
| `LABEVENTS` | `LAB//{itemid}//{unit}` | `numeric_value`, `text_value` and `flag` are kept. |
|
|
85
|
+
| `CHARTEVENTS` | `CHART//{itemid}//{unit}` | |
|
|
86
|
+
| `INPUTEVENTS` | `FLUID_INPUT//{unit}` | PIC records only fluid intake amounts here, with no item ID. |
|
|
87
|
+
| `OUTPUTEVENTS` | `OUTPUT//{itemid}//{unit}` | |
|
|
88
|
+
| `PRESCRIPTIONS` | `MEDICATION//START//{drug}//{dose unit}`, `MEDICATION//END//{drug}` | Uses the English drug name. The dose is the `numeric_value`. |
|
|
89
|
+
| `MICROBIOLOGYEVENTS` | `MICROBIOLOGY//SPECIMEN//{spec_itemid}`, `MICROBIOLOGY//ORGANISM//{org_itemid}`, `MICROBIOLOGY//SUSCEPTIBILITY//{ab_itemid}//{S/I/R…}` | The MIC dilution is the `numeric_value`. Organism and specimen IDs are kept as columns. |
|
|
90
|
+
| `OR_EXAM_REPORTS` | `EXAM//{type}//{item}` | Uses the exam time, or the report time when the exam time is missing. |
|
|
91
|
+
| `EMR_SYMPTOMS` | `SYMPTOM//{symptom}//{+/-}` | Symptoms extracted by NLP from clinical notes. |
|
|
92
|
+
| `SURGERY_INFO` | `SURGERY//START//{name}`, `SURGERY//END//{name}`, `ANESTHESIA//START//{method}`, `ANESTHESIA//END//{method}` | |
|
|
93
|
+
| `SURGERY_VITAL_SIGNS` | `SURGERY_VITAL//{itemid}` | Intra-operative monitoring. Units are in the code metadata (`unit`). |
|
|
94
|
+
|
|
95
|
+
A missing unit or susceptibility interpretation becomes `UNK` in the code. A row with a missing
|
|
96
|
+
identifying component (for example, no organism) produces no event for that component.
|
|
97
|
+
|
|
98
|
+
### Code metadata
|
|
99
|
+
|
|
100
|
+
`metadata/codes.parquet` links codes to PIC's dictionary tables:
|
|
101
|
+
|
|
102
|
+
- **`description`**: the English label from `D_ITEMS`, `D_LABITEMS` or `D_ICD_DIAGNOSES`.
|
|
103
|
+
- **`parent_codes`**: `LOINC/…` for lab items with a well-formed LOINC code, and `ICD10/…` for
|
|
104
|
+
diagnoses whose mapped code is a valid 3- or 4-character WHO ICD-10 code. PIC's own extended codes
|
|
105
|
+
(7+ characters) have no parent.
|
|
106
|
+
- **Chinese source terms**: kept in their own columns (`title_cn`, `label_cn`, `drug_name_cn`,
|
|
107
|
+
`org_name_cn`, `spec_type_desc_cn`, `ab_name_cn`, `symptom_name_cn`). Codes are always built from the
|
|
108
|
+
English or ID columns.
|
|
109
|
+
- **Extra dictionary fields**: `fluid`, `category` and `unit`.
|
|
110
|
+
|
|
111
|
+
### Data notes
|
|
112
|
+
|
|
113
|
+
- **Shifted dates.** Dates are shifted 50–100 years into the future per patient. Years between 2049 and
|
|
114
|
+
2119 are expected, and intervals within a patient are preserved.
|
|
115
|
+
- **Release layout.** The v1.1.0 release ships the current tables under `V1.1.0/`. The gzipped files at
|
|
116
|
+
the release root are the superseded v1.0.0 tables, and this ETL does not read them.
|
|
117
|
+
- **Swapped column names.** In `PRESCRIPTIONS`, the CSV header `DRUG_NAME` holds the Chinese name and
|
|
118
|
+
`DRUG_NAME_EN` the English one. The bundled SQL scripts name these columns differently.
|
|
119
|
+
- **Many-to-many microbiology IDs.** Item IDs and their Chinese names do not map one-to-one, so codes
|
|
120
|
+
use the item IDs. Expect MEDS-Extract warnings that some codes have several `*_cn` values; those
|
|
121
|
+
values are aggregated into lists.
|
|
122
|
+
- **Uninformative columns.** `LANGUAGE`, `RELIGION` and `MARITAL_STATUS` hold a single value across the
|
|
123
|
+
release and are not extracted.
|
|
124
|
+
- **Local diagnosis codes.** Diagnoses use PIC's extended Chinese ICD-10 codes (`ICD10_CN`), so they do
|
|
125
|
+
not yet line up with standard ICD-10 codes in other datasets. Valid standard codes are attached as
|
|
126
|
+
`parent_codes`. Mapping equivalent codes to standard ICD-10 needs a MEDS-Extract feature; see
|
|
127
|
+
[#1](https://github.com/rvandewater/PICdb_MEDS/issues/1).
|
|
128
|
+
|
|
129
|
+
## Testing
|
|
130
|
+
|
|
131
|
+
PIC has no open demo release. Instead, `tests/e2e_synthetic_test.py` writes a few fully synthetic rows
|
|
132
|
+
with PIC's exact headers and layout, runs `meds-extract-run spec=PICdb` over them, and checks the
|
|
133
|
+
events, times, values and code metadata it produces.
|
|
134
|
+
|
|
135
|
+
```bash
|
|
136
|
+
uv sync --group dev
|
|
137
|
+
uv run pytest tests/ -v
|
|
138
|
+
uv run pre-commit run --all-files
|
|
139
|
+
```
|
|
140
|
+
|
|
141
|
+
## Citation
|
|
142
|
+
|
|
143
|
+
If you use this ETL, please cite the PIC database and its descriptor paper:
|
|
144
|
+
|
|
145
|
+
- Li, H., Zeng, X., & Yu, G. (2020). *Paediatric Intensive Care database* (version 1.1.0). PhysioNet.
|
|
146
|
+
https://doi.org/10.13026/32x9-wv38
|
|
147
|
+
- Zeng, X., Yu, G., Lu, Y. et al. PIC, a paediatric-specific intensive care database. *Sci Data* 7, 14
|
|
148
|
+
(2020). https://doi.org/10.1038/s41597-020-0355-4
|
|
149
|
+
|
|
150
|
+
See also [`CITATION.cff`](CITATION.cff).
|
|
@@ -0,0 +1,68 @@
|
|
|
1
|
+
[build-system]
|
|
2
|
+
requires = ["setuptools>=64", "wheel"]
|
|
3
|
+
build-backend = "setuptools.build_meta"
|
|
4
|
+
|
|
5
|
+
[project]
|
|
6
|
+
name = "PICdb-MEDS"
|
|
7
|
+
# Bump with `uv version --bump patch|minor|major`, then tag the commit with the same version.
|
|
8
|
+
version = "0.0.1"
|
|
9
|
+
authors = [
|
|
10
|
+
{ name="Robin van de Water", email="rpv2114@cumc.columbia.edu" },
|
|
11
|
+
]
|
|
12
|
+
description = "An ETL pipeline to extract the Paediatric Intensive Care (PIC) database into the MEDS format."
|
|
13
|
+
readme = "README.md"
|
|
14
|
+
# Mirrors MEDS-Extract 0.7's own interpreter range; a looser range here admits interpreters it
|
|
15
|
+
# cannot be installed on.
|
|
16
|
+
requires-python = ">=3.11.4,<3.14"
|
|
17
|
+
classifiers = [
|
|
18
|
+
"Programming Language :: Python :: 3",
|
|
19
|
+
"License :: OSI Approved :: MIT License",
|
|
20
|
+
"Operating System :: OS Independent",
|
|
21
|
+
]
|
|
22
|
+
# This package ships no Python: the whole ETL is `event_configs.yaml`, run by MEDS-Extract's own
|
|
23
|
+
# `meds-extract-run`. The `[download]` extra adds `meds-extract-download`, which the runner spawns to
|
|
24
|
+
# stage the raw data from PhysioNet. MEDS-transforms, polars and hydra arrive through it at the
|
|
25
|
+
# versions it fences.
|
|
26
|
+
#
|
|
27
|
+
# The floor is 0.7.0: this config uses `_self` metadata blocks, aggregated joins and regex
|
|
28
|
+
# conditionals in `parent_codes`, none of which exist before 0.7.
|
|
29
|
+
dependencies = [
|
|
30
|
+
"meds-extract[download]>=0.7.0,<0.8",
|
|
31
|
+
]
|
|
32
|
+
|
|
33
|
+
[dependency-groups]
|
|
34
|
+
dev = [
|
|
35
|
+
"pre-commit~=3.8",
|
|
36
|
+
"ruff>=0.11,<1.0",
|
|
37
|
+
"pytest~=9.0",
|
|
38
|
+
"shfmt-py==3.12.0.2",
|
|
39
|
+
]
|
|
40
|
+
|
|
41
|
+
[project.entry-points."MEDS_extract.pipelines"]
|
|
42
|
+
PICdb = "PICdb_MEDS:event_configs.yaml"
|
|
43
|
+
|
|
44
|
+
[project.urls]
|
|
45
|
+
Homepage = "https://github.com/rvandewater/PICdb_MEDS"
|
|
46
|
+
Issues = "https://github.com/rvandewater/PICdb_MEDS/issues"
|
|
47
|
+
|
|
48
|
+
[tool.ruff]
|
|
49
|
+
target-version = "py311"
|
|
50
|
+
line-length = 110
|
|
51
|
+
|
|
52
|
+
[tool.ruff.lint]
|
|
53
|
+
select = [
|
|
54
|
+
"E", # pycodestyle
|
|
55
|
+
"F", # pyflakes
|
|
56
|
+
"I", # isort
|
|
57
|
+
"UP", # pyupgrade
|
|
58
|
+
"B", # flake8-bugbear
|
|
59
|
+
"C4", # flake8-comprehensions
|
|
60
|
+
"RUF", # Ruff-specific rules
|
|
61
|
+
"SIM", # flake8-simplify
|
|
62
|
+
]
|
|
63
|
+
ignore = [
|
|
64
|
+
"SIM108", # ternary operator suggestions
|
|
65
|
+
]
|
|
66
|
+
|
|
67
|
+
[tool.ruff.lint.per-file-ignores]
|
|
68
|
+
"__init__.py" = ["F401"]
|
|
@@ -0,0 +1,12 @@
|
|
|
1
|
+
"""The PIC (Paediatric Intensive Care database) MEDS ETL — a configuration package, not a library.
|
|
2
|
+
|
|
3
|
+
This distribution ships no code. Its entire content is ``PICdb_MEDS/event_configs.yaml``, a MESSY
|
|
4
|
+
(MEDS-Extract Specification Syntax YAML) document declaring where the raw PIC data comes from and how
|
|
5
|
+
every event is extracted from it. ``pyproject.toml`` registers that file with MEDS-Extract's
|
|
6
|
+
``MEDS_extract.pipelines`` entry-point group under the name ``PICdb``, so the ETL runs as::
|
|
7
|
+
|
|
8
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR
|
|
9
|
+
|
|
10
|
+
This package exists so the YAML is addressable as package data —
|
|
11
|
+
``importlib.resources.files("PICdb_MEDS")`` — from an installed wheel. Nothing here is importable API.
|
|
12
|
+
"""
|
|
@@ -0,0 +1,166 @@
|
|
|
1
|
+
Metadata-Version: 2.4
|
|
2
|
+
Name: PICdb-MEDS
|
|
3
|
+
Version: 0.0.1
|
|
4
|
+
Summary: An ETL pipeline to extract the Paediatric Intensive Care (PIC) database into the MEDS format.
|
|
5
|
+
Author-email: Robin van de Water <rpv2114@cumc.columbia.edu>
|
|
6
|
+
Project-URL: Homepage, https://github.com/rvandewater/PICdb_MEDS
|
|
7
|
+
Project-URL: Issues, https://github.com/rvandewater/PICdb_MEDS/issues
|
|
8
|
+
Classifier: Programming Language :: Python :: 3
|
|
9
|
+
Classifier: License :: OSI Approved :: MIT License
|
|
10
|
+
Classifier: Operating System :: OS Independent
|
|
11
|
+
Requires-Python: <3.14,>=3.11.4
|
|
12
|
+
Description-Content-Type: text/markdown
|
|
13
|
+
License-File: LICENSE
|
|
14
|
+
Requires-Dist: meds-extract[download]<0.8,>=0.7.0
|
|
15
|
+
Dynamic: license-file
|
|
16
|
+
|
|
17
|
+
# PICdb MEDS Extraction ETL
|
|
18
|
+
|
|
19
|
+
<!-- Placeholders until available: the PyPI badges resolve once PICdb-MEDS is published; replace
|
|
20
|
+
CODECOV_TOKEN_PLACEHOLDER with the repo's Codecov badge token and zenodo.XXXXXXX with its Zenodo DOI. -->
|
|
21
|
+
|
|
22
|
+
[](https://pypi.org/project/PICdb-MEDS/)
|
|
23
|
+
[](https://codecov.io/gh/rvandewater/PICdb_MEDS)
|
|
24
|
+
[](https://github.com/rvandewater/PICdb_MEDS/actions/workflows/synthetic-etl.yaml)
|
|
25
|
+
[](https://github.com/rvandewater/PICdb_MEDS/actions/workflows/code-quality-main.yaml)
|
|
26
|
+
|
|
27
|
+
|
|
28
|
+
[](https://github.com/rvandewater/PICdb_MEDS#license)
|
|
29
|
+
[](https://github.com/rvandewater/PICdb_MEDS/pulls)
|
|
30
|
+
[](https://github.com/rvandewater/PICdb_MEDS/graphs/contributors)
|
|
31
|
+
[](https://doi.org/10.5281/zenodo.XXXXXXX)
|
|
32
|
+
|
|
33
|
+
This pipeline extracts the [Paediatric Intensive Care (PIC) database v1.1.0](https://physionet.org/content/picdb/1.1.0/)
|
|
34
|
+
from PhysioNet into the [MEDS format](https://medical-event-data-standard.github.io/). PIC holds 12,881
|
|
35
|
+
neonatal and paediatric patients and 13,449 hospital admissions from the Children's Hospital of Zhejiang
|
|
36
|
+
University School of Medicine (2010–2018). It was built from the
|
|
37
|
+
[ETL_MEDS_Template](https://github.com/Medical-Event-Data-Standard/ETL_MEDS_Template) and uses the
|
|
38
|
+
config-only layout introduced by [MEDS-Extract](https://github.com/mmcdermott/MEDS_extract) 0.7.
|
|
39
|
+
|
|
40
|
+
## Usage
|
|
41
|
+
|
|
42
|
+
PIC is a credentialed dataset. You need a PhysioNet account that has signed the PIC data use agreement.
|
|
43
|
+
|
|
44
|
+
```bash
|
|
45
|
+
pip install git+https://github.com/rvandewater/PICdb_MEDS.git
|
|
46
|
+
export DATASET_DOWNLOAD_USERNAME=$PHYSIONET_USERNAME
|
|
47
|
+
export DATASET_DOWNLOAD_PASSWORD=$PHYSIONET_PASSWORD
|
|
48
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR
|
|
49
|
+
```
|
|
50
|
+
|
|
51
|
+
When you run this, the program will:
|
|
52
|
+
|
|
53
|
+
1. Download the v1.1.0 tables (`V1.1.0/*.csv`, about 1.2 GB) from PhysioNet into
|
|
54
|
+
`$MEDS_OUTPUT_DIR/.meds_extract_run/raw_input`. Files that already exist and match their checksums
|
|
55
|
+
are skipped, so an interrupted download resumes. Pass `download_dest_dir=$RAW_INPUT_DIR` to keep the
|
|
56
|
+
raw data somewhere durable and reuse it across runs.
|
|
57
|
+
2. Build the MEDS cohort from those files and write it to `$MEDS_OUTPUT_DIR` (`data/` and `metadata/`).
|
|
58
|
+
All transformations, joins and metadata extraction are declared in
|
|
59
|
+
[`src/PICdb_MEDS/event_configs.yaml`](src/PICdb_MEDS/event_configs.yaml).
|
|
60
|
+
|
|
61
|
+
If you already have the PhysioNet release on disk, skip the download and point `input_dir` at the
|
|
62
|
+
release root, the directory that contains `V1.1.0/`:
|
|
63
|
+
|
|
64
|
+
```bash
|
|
65
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR \
|
|
66
|
+
do_download=false input_dir=/path/to/paediatric-intensive-care-database-1.1.0
|
|
67
|
+
```
|
|
68
|
+
|
|
69
|
+
To parallelize the stages, pass a runner file:
|
|
70
|
+
|
|
71
|
+
```bash
|
|
72
|
+
printf 'parallelize:\n n_workers: 8\n launcher: joblib\n' >runner.yaml
|
|
73
|
+
meds-extract-run spec=PICdb output_dir=$MEDS_OUTPUT_DIR stage_runner_fp=runner.yaml
|
|
74
|
+
```
|
|
75
|
+
|
|
76
|
+
Run `meds-extract-run --help` for every option.
|
|
77
|
+
|
|
78
|
+
## How this ETL is defined
|
|
79
|
+
|
|
80
|
+
The package contains no Python. The whole ETL is one MESSY (MEDS-Extract Specification Syntax YAML)
|
|
81
|
+
file, [`src/PICdb_MEDS/event_configs.yaml`](src/PICdb_MEDS/event_configs.yaml). It declares:
|
|
82
|
+
|
|
83
|
+
- **`sources:`**: where the raw data lives (the PhysioNet release, restricted to `V1.1.0/*.csv`).
|
|
84
|
+
- **`etl:`**: the dataset name and sharding options.
|
|
85
|
+
- **One block per raw table**: how each row becomes one or more MEDS events, and which dictionary
|
|
86
|
+
tables describe the resulting codes.
|
|
87
|
+
|
|
88
|
+
`pyproject.toml` registers the file with MEDS-Extract's `MEDS_extract.pipelines` entry-point group under
|
|
89
|
+
the name `PICdb`. That registration is what makes `spec=PICdb` work. No pre-MEDS step is needed: every
|
|
90
|
+
PIC table already carries `SUBJECT_ID`, and every timestamp has the same format.
|
|
91
|
+
|
|
92
|
+
## Events
|
|
93
|
+
|
|
94
|
+
| Table | Event codes | Notes |
|
|
95
|
+
| --------------------- | -------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------- |
|
|
96
|
+
| `PATIENTS` | `GENDER//{gender}`, `MEDS_BIRTH`, `MEDS_DEATH` | Death time is the earliest `ADMISSIONS.DEATHTIME`, falling back to `DOD`. |
|
|
97
|
+
| `ADMISSIONS` | `HOSPITAL_ADMISSION//{dept}`, `HOSPITAL_DISCHARGE//{dept}`, `INSURANCE//…`, `ETHNICITY//…`, `ADMISSION_DIAGNOSIS//ICD10_CN//{code}` | ED times equal admit/discharge times in every row, so there are no separate ED events. |
|
|
98
|
+
| `ICUSTAYS` | `ICU_ADMISSION//{careunit}`, `ICU_DISCHARGE//{careunit}` | |
|
|
99
|
+
| `DIAGNOSES_ICD` | `DIAGNOSIS//ICD10_CN//{code}` | Placed at hospital discharge. `seq_num` and `diag_category` are kept as columns. |
|
|
100
|
+
| `LABEVENTS` | `LAB//{itemid}//{unit}` | `numeric_value`, `text_value` and `flag` are kept. |
|
|
101
|
+
| `CHARTEVENTS` | `CHART//{itemid}//{unit}` | |
|
|
102
|
+
| `INPUTEVENTS` | `FLUID_INPUT//{unit}` | PIC records only fluid intake amounts here, with no item ID. |
|
|
103
|
+
| `OUTPUTEVENTS` | `OUTPUT//{itemid}//{unit}` | |
|
|
104
|
+
| `PRESCRIPTIONS` | `MEDICATION//START//{drug}//{dose unit}`, `MEDICATION//END//{drug}` | Uses the English drug name. The dose is the `numeric_value`. |
|
|
105
|
+
| `MICROBIOLOGYEVENTS` | `MICROBIOLOGY//SPECIMEN//{spec_itemid}`, `MICROBIOLOGY//ORGANISM//{org_itemid}`, `MICROBIOLOGY//SUSCEPTIBILITY//{ab_itemid}//{S/I/R…}` | The MIC dilution is the `numeric_value`. Organism and specimen IDs are kept as columns. |
|
|
106
|
+
| `OR_EXAM_REPORTS` | `EXAM//{type}//{item}` | Uses the exam time, or the report time when the exam time is missing. |
|
|
107
|
+
| `EMR_SYMPTOMS` | `SYMPTOM//{symptom}//{+/-}` | Symptoms extracted by NLP from clinical notes. |
|
|
108
|
+
| `SURGERY_INFO` | `SURGERY//START//{name}`, `SURGERY//END//{name}`, `ANESTHESIA//START//{method}`, `ANESTHESIA//END//{method}` | |
|
|
109
|
+
| `SURGERY_VITAL_SIGNS` | `SURGERY_VITAL//{itemid}` | Intra-operative monitoring. Units are in the code metadata (`unit`). |
|
|
110
|
+
|
|
111
|
+
A missing unit or susceptibility interpretation becomes `UNK` in the code. A row with a missing
|
|
112
|
+
identifying component (for example, no organism) produces no event for that component.
|
|
113
|
+
|
|
114
|
+
### Code metadata
|
|
115
|
+
|
|
116
|
+
`metadata/codes.parquet` links codes to PIC's dictionary tables:
|
|
117
|
+
|
|
118
|
+
- **`description`**: the English label from `D_ITEMS`, `D_LABITEMS` or `D_ICD_DIAGNOSES`.
|
|
119
|
+
- **`parent_codes`**: `LOINC/…` for lab items with a well-formed LOINC code, and `ICD10/…` for
|
|
120
|
+
diagnoses whose mapped code is a valid 3- or 4-character WHO ICD-10 code. PIC's own extended codes
|
|
121
|
+
(7+ characters) have no parent.
|
|
122
|
+
- **Chinese source terms**: kept in their own columns (`title_cn`, `label_cn`, `drug_name_cn`,
|
|
123
|
+
`org_name_cn`, `spec_type_desc_cn`, `ab_name_cn`, `symptom_name_cn`). Codes are always built from the
|
|
124
|
+
English or ID columns.
|
|
125
|
+
- **Extra dictionary fields**: `fluid`, `category` and `unit`.
|
|
126
|
+
|
|
127
|
+
### Data notes
|
|
128
|
+
|
|
129
|
+
- **Shifted dates.** Dates are shifted 50–100 years into the future per patient. Years between 2049 and
|
|
130
|
+
2119 are expected, and intervals within a patient are preserved.
|
|
131
|
+
- **Release layout.** The v1.1.0 release ships the current tables under `V1.1.0/`. The gzipped files at
|
|
132
|
+
the release root are the superseded v1.0.0 tables, and this ETL does not read them.
|
|
133
|
+
- **Swapped column names.** In `PRESCRIPTIONS`, the CSV header `DRUG_NAME` holds the Chinese name and
|
|
134
|
+
`DRUG_NAME_EN` the English one. The bundled SQL scripts name these columns differently.
|
|
135
|
+
- **Many-to-many microbiology IDs.** Item IDs and their Chinese names do not map one-to-one, so codes
|
|
136
|
+
use the item IDs. Expect MEDS-Extract warnings that some codes have several `*_cn` values; those
|
|
137
|
+
values are aggregated into lists.
|
|
138
|
+
- **Uninformative columns.** `LANGUAGE`, `RELIGION` and `MARITAL_STATUS` hold a single value across the
|
|
139
|
+
release and are not extracted.
|
|
140
|
+
- **Local diagnosis codes.** Diagnoses use PIC's extended Chinese ICD-10 codes (`ICD10_CN`), so they do
|
|
141
|
+
not yet line up with standard ICD-10 codes in other datasets. Valid standard codes are attached as
|
|
142
|
+
`parent_codes`. Mapping equivalent codes to standard ICD-10 needs a MEDS-Extract feature; see
|
|
143
|
+
[#1](https://github.com/rvandewater/PICdb_MEDS/issues/1).
|
|
144
|
+
|
|
145
|
+
## Testing
|
|
146
|
+
|
|
147
|
+
PIC has no open demo release. Instead, `tests/e2e_synthetic_test.py` writes a few fully synthetic rows
|
|
148
|
+
with PIC's exact headers and layout, runs `meds-extract-run spec=PICdb` over them, and checks the
|
|
149
|
+
events, times, values and code metadata it produces.
|
|
150
|
+
|
|
151
|
+
```bash
|
|
152
|
+
uv sync --group dev
|
|
153
|
+
uv run pytest tests/ -v
|
|
154
|
+
uv run pre-commit run --all-files
|
|
155
|
+
```
|
|
156
|
+
|
|
157
|
+
## Citation
|
|
158
|
+
|
|
159
|
+
If you use this ETL, please cite the PIC database and its descriptor paper:
|
|
160
|
+
|
|
161
|
+
- Li, H., Zeng, X., & Yu, G. (2020). *Paediatric Intensive Care database* (version 1.1.0). PhysioNet.
|
|
162
|
+
https://doi.org/10.13026/32x9-wv38
|
|
163
|
+
- Zeng, X., Yu, G., Lu, Y. et al. PIC, a paediatric-specific intensive care database. *Sci Data* 7, 14
|
|
164
|
+
(2020). https://doi.org/10.1038/s41597-020-0355-4
|
|
165
|
+
|
|
166
|
+
See also [`CITATION.cff`](CITATION.cff).
|
|
@@ -0,0 +1,10 @@
|
|
|
1
|
+
LICENSE
|
|
2
|
+
README.md
|
|
3
|
+
pyproject.toml
|
|
4
|
+
src/PICdb_MEDS/__init__.py
|
|
5
|
+
src/PICdb_MEDS.egg-info/PKG-INFO
|
|
6
|
+
src/PICdb_MEDS.egg-info/SOURCES.txt
|
|
7
|
+
src/PICdb_MEDS.egg-info/dependency_links.txt
|
|
8
|
+
src/PICdb_MEDS.egg-info/entry_points.txt
|
|
9
|
+
src/PICdb_MEDS.egg-info/requires.txt
|
|
10
|
+
src/PICdb_MEDS.egg-info/top_level.txt
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
meds-extract[download]<0.8,>=0.7.0
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
PICdb_MEDS
|