API reference

This page provides an auto-generated summary of mesmers’ API.

Top-level functions

set_options(**kwargs)	Set options for mesmer in a controlled context.
get_options()	Get options for mesmer.

Statistical functions

Linear regression

LinearRegression()	Ordinary least squares Linear Regression for xr.DataArray objects.
from_params(params)	initialize LinearRegression class using parameters
fit(predictors, target, dim[, weights, ...])	Fit a linear model
predict(predictors, *[, exclude, only])	Predict using the linear model.
residuals(predictors, target)	Calculate the residuals of the fitted linear model
to_netcdf(filename, **kwargs)	save params to a netCDF file
from_netcdf(filename, **kwargs)	read params from a netCDF file

Auto regression

select_ar_order_scen_ens(*objs, dim, ...[, ic])	Select the order of an autoregressive process and potentially calculate the median over ensemble members and scenarios
fit_auto_regression_scen_ens(*objs, dim, ...)	fit an auto regression and potentially calculate the mean over ensemble members and scenarios
select_ar_order(data, dim, *, maxlag[, ic])	Select the order of an autoregressive process
fit_auto_regression(data, dim, *, lags)	fit an auto regression
fit_auto_regression_monthly(monthly_data[, ...])	fit a cyclo-stationary auto-regressive process of lag one (AR(1)) on monthly data.
draw_auto_regression_uncorrelated(ar_params, ...)	draw time series of an auto regression process
draw_auto_regression_correlated(ar_params, ...)	draw time series of an auto regression process with spatially-correlated innovations
draw_auto_regression_monthly(ar_params, ...)	draw time series of a cyclo-stationary auto-regressive process of lag one (AR(1)) using individual parameters for each month including spatially-correlated innovations.

Harmonic Model

fit_harmonic_model(yearly_predictor, ...[, ...])	fit harmonic model i.e. a Fourier Series to every gridcell using BIC score to select the order and least squares to fit the coefficients for each order.
predict_harmonic_model(yearly_predictor, ...)	construct a Fourier Series from yearly predictors with fitted coeffs.

Power Transformer

YeoJohnsonTransformer(name)
lambda_function(coeffs, local_yearly_T)	function defining the covariate function to calculate lambda depending on the local yearly values.
get_lambdas_from_covariates(lambda_coeffs, ...)	function that relates fitted coefficients and the yearly predictor to the lambdas.
fit(yearly_pred, monthly_residuals[, time_dim])	estimate the optimal coefficients for the parameters \(\lambda\) for each gridcell, to normalize monthly residuals conditional on yearly predictor.
transform(yearly_pred, monthly_residuals, ...)	transform monthly_residuals following Yeo-Johnson transformer with parameters \(\lambda\), fit with fit_yeo_johnson_transform.
inverse_transform(yearly_pred, ...[, time_dim])	apply the inverse power transformation using the fitted lambdas.
constant_lambda_function(coeffs, local_yearly_T)	Use a constant for lambda.
logistic_lambda_function(coeffs, local_yearly_T)	Use logistic function to calculate lambda depending on the local yearly values.

Localized covariance

adjust_covariance_ar1(covariance, ar_coefs)	adjust localized empirical covariance matrix for autoregressive process of order one
find_localized_empirical_covariance(data, ...)	determine localized empirical covariance by cross validation
find_localized_empirical_covariance_monthly(...)	determine localized empirical covariance by cross validation for each month.

Smoothing

lowess(data, dim, *[, combine_dim, n_steps, ...])

LOWESS (Locally Weighted Scatterplot Smoothing) for xarray objects

Gaspari-Cohn correlation matrix

gaspari_cohn_correlation_matrices(geodist, ...)	Gaspari-Cohn correlation matrices for a range of localisation radii
gaspari_cohn(r)	smooth, exponentially decaying Gaspari-Cohn correlation function

Conditional distribution

Define covariance structure of conditional distribution

Expression(expr, expr_name[, ...])
evaluate_params(coefficients_values, ...[, ...])	Evaluates the parameters for the provided predictors and coefficients

Fit conditional distribution

ConditionalDistribution(expression, *[, ...])
find_first_guess(predictors, target, weights)	Find a first guess for all coefficients of a conditional distribution
fit(predictors, target, weights, first_guess, *)	fit conditional distribution over all gridpoints.
compute_quality_scores(predictors, target, ...)	Compute scores for fit coefficients.
coefficients	The coefficients of this conditional distribution.
from_netcdf(filename, **kwargs)	read coefficients from a netCDF file with default solver options
to_netcdf(filename, **kwargs)	save coefficients dataset to a netCDF file

MinimizeOptions([method, tol, options])

Transform conditional distribution

ProbabilityIntegralTransform(distrib_orig, ...)
transform(data, target_name[, preds_orig, ...])	Probability integral transform data given coefficients for the expression of a conditional distribution.

Data handling

Example and test data

cmip6_ng_path(*[, relative])

path of the cmip6_ng example data

Grid manipulation

wrap_to_180(obj[, lon_name])	wrap array with longitude to [-180..180)
wrap_to_360(obj[, lon_name])	wrap array with longitude to [0..360)
stack_lat_lon(data, *[, x_dim, y_dim, ...])	Stack a regular lat-lon grid to a 1D (unstructured) grid
unstack_lat_lon_and_align(data, coords_orig, *)	unstack an 1D grid to a regular lat-lon grid and align with original coords
unstack_lat_lon(data, *[, x_dim, y_dim, ...])	unstack an 1D grid to a regular lat-lon grid but do not align
align_to_coords(data, coords_orig)	align an unstacked lat-lon grid with its original coords

Masking regions

mask_ocean_fraction(data, threshold, *[, ...])	mask out ocean using fractional overlap
mask_ocean(data, *[, x_coords, y_coords])	mask out ocean
mask_antarctica(data, *[, y_coords])	mask out ocean

Weighted operations: calculate global mean

global_mean(data[, weights, x_dim, y_dim])	calculate global weighted mean
lat_weights(data[, y_dim])	area weights based on the cosine of the latitude
weighted_mean(data, weights, /, *[, dims])	weighted mean - convenience function which ignores data_vars missing dims
equal_scenario_weights_from_datatree(dt[, ...])	Create a DataTree isomorphic to dt, holding the weights for each scenario to weight the ensemble members of each scenario such that each scenario contributes equally to some fitting procedure.
get_weights_density(pred_data)	generate inverse data-density weights

DataTree heplers

pool_scen_ens(dt, *[, member_dim, time_dim, ...])	prepare data for statistical functions
broadcast_and_pool_scen_ens(predictors, target)	prepare predictors, target, and weights for statistical functions
merge(objects[, compat, join, fill_value, ...])	Merge the datasets of each node of isomorphic DataTree objects together.
collapse_datatree_into_dataset(dt, *, dim, ...)	Take a DataTree and collapse all its subtrees into a single xr.Dataset along dim.
map_over_datasets(func, *args[, kwargs])	Applies a function to every dataset in one or more DataTree objects with the same structure (ie.., that are isomorphic), returning new trees which store the results.

Geospatial

geodist_exact(lon, lat, *[, equal_dim_suffixes])

exact great circle distance based on WSG 84

Anomalies

calc_anomaly(dt, reference_period, *[, ...])

subtract mean over the reference period

Resampling

upsample_yearly_data(yearly_data, monthly_time)

Upsample yearly data to monthly resolution by repeating yearly values.

Emulator functions

Volcanic influence

fit_volcanic_influence(tas_residuals[, ...])	estimate volcanic influence on temperature residuals using aerosol optical depth observations as proxy
superimpose_volcanic_influence(...[, ...])	superimpose volcanic influence on smooth temperature anomalies using aerosol optical depth observations as proxy
load_stratospheric_aerosol_optical_depth_obs(*)	load stratospheric aerosol optical depth data - a proxy for volcanic activity