Skip to main content

Electronic Medical Record Machine Learning Utilities for using Models with health data.

https://badge.fury.io/py/ehrml.svg

Configuration:

These utilities are very dependent on a particular configuration format. In python, it is a list of dicts, where each dict represents configuration for a particular field. The keys in this dict are as follows:

field

Description

index

The index to write the value (or in the case of any one-hot field, to start writing values) in the numpy array.

missing_flag_index

The index to write a one to if the data is missing (pre-imputation) in the numpy array. Do not set if no such missing data flag is desired for this field.

rwb_src

The value used to represent the value for all observation lists. Also, the field name associated with the “flat” data source, without any time suffix.

api_parent

The key in the layered data which contains relevant data for this field.

api_time_src

Which field in the layered data contains a reference to datetime for this observation.

api_src

Regarding the layered data, either the direct access field for each item in the list under api_parent, or the desired value of api_by.

api_by

If a field in layered data is not direct access, this is the field under api_parent which contains the name matching api_src. Do not set for direct access values.

api_from

If a field in layered data is not direct access, this is the field under api_parent which contains the value. Do not set for direct access values.

transformation

The name of a transformation or encoding to be executed on this field.

one_hot_vals

An array of values corresponding to a one hot encoding for this field. Different for each encoding, unused for numerical transformations.

mean

For numeric transformations, the precomputed mean.

std

For numeric transformations, the precomputed standard deviation.

min

For numeric transformations, replace any value lower than this value with this value. Also used in some transformations.

max

For numeric transformations, replace any value higher than this value with this value. Also used in some transformations.

Documentation

This tool uses a few different input and output structures in order to facilitate computation and analysis. Descriptions of these formats, along with descriptions of the methods and their inputs are in the python docstrings for these methods.

Note

At this point, the utilities here may be very specific to a particular kind of EHR and model.

Release files for ehrml 1.0.3

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for ehrml 1.0.3
File Size Uploaded
ehrml-1.0.3.tar.gz 20.3 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for ehrml 1.0.3
File Interpreter ABI Platform
ehrml-1.0.3-py3-none-any.whl Python 3 none any Details

Total release size:40.8 kB

Release files / ehrml-1.0.3.tar.gz

Download URL ehrml-1.0.3.tar.gz
Size 20.3 kB
Tags Source
SHA-256 checksum
How to use checksums
33c3a485ced41cd1efc16fa508245550a9b514a2ad5406289da8f7732e87f9c4
BLAKE2b-256 checksum
How to use checksums
37206b588817442dbee0e34dc15f040099fc0ec72ca06dffe920320166edf993
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/4.0.1 CPython/3.9.13

Release files / ehrml-1.0.3-py3-none-any.whl

Download URL ehrml-1.0.3-py3-none-any.whl
Size 20.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
12ccf056115be2256cf58b9c7d80326606a141fdbcedbfd254650a28e9501810
BLAKE2b-256 checksum
How to use checksums
891b208f0f8f64f99fc4f023b483478fede68c105081ffe59dd232b6569de173
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/4.0.1 CPython/3.9.13

Release history Release notifications | RSS feed

This release

1.0.3 This release

2 release files

1.0.2

2 release files

1.0.1

2 release files

1.0

1 release file

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page