Tools to analyse climate data for machine learning and event analysis
Project description
climpy
Climpy is working to help climate researchers to analyse climate data, write in formats ready to be used with machine learning models and analyse the accuracy of model predictions
https://github.com/climai/climpy/actions/workflows/python-app.yml/badge.svg
The package is divided into three parts
- Transform: The
transform
module transforms by applying different conditions on your dataset. The class diagram below will detail on the application of the module.
---
Transform
---
classDiagram
Hazard <|-- Criterion
class Condition{
args
returns_event
func()
}
class Criterion{
sequence
apply_conditions()
valid_conditions(sequence) bool
}
class Hazard{
event_locations
n_events
apply_conditions()
valid_conditions()
get_event(n) Event
all_events() EventList
}
class DataArray{
xr.DataArray variables
xr.DataArray functions()
}
class LinkDataHazard{
on_events()
get_values()
}
class Event{
data
location
start_time
end_time
r
tau
set_intensity()
}
class EventList{
event_list
}
DataArray -- LinkDataHazard
LinkDataHazard -- Hazard
Hazard *-- EventList
EventList *--Event
Condition .. Criterion
Condition .. Hazard
- ml_data: The
ml_data
module creates/writes data to be used conveniently for different kinds of machine learning models. The class diagram below will detail on the application of the module.
---
Data ML
---
classDiagram
List *-- DataArray
class X{
values
meta
}
class Y{
values
meta
}
class MLData{
X
Y
tvt_split()
}
class DataArray{
xr.DataArray variables
xr.DataArray functions()
}
class Split
- Metrics: The
metrics
module can be applied to observed and simulated variables. This would include exhaustive set of different metrics that can be used on climate related data
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
climpy-0.0.4.tar.gz
(8.2 kB
view hashes)
Built Distribution
climpy-0.0.4-py3-none-any.whl
(10.4 kB
view hashes)