MLJ Cheatsheet

Starting an interactive MLJ session

julia> using MLJ

julia> MLJ_VERSION # version of MLJ for this cheatsheet
"0.7.0"

Model search and code loading

info("PCA") retrieves registry metadata for the model called "PCA"

info("RidgeRegressor", pkg="MultivariateStats") retrieves metadata for "RidgeRegresssor", which is provided by multiple packages

models() lists metadata of every registered model.

models(x -> x.is_supervised && x.is_pure_julia) lists all supervised models written in pure julia.

experimental: models(matching(X)) lists all unsupervised models compatible with input X.

experimental! models(matching(X, y)) lists all supervised modesl compatible with input/target X/y.

experimental! With additional conditions:

models() do model
    matching(model, X, y)) &&
    model.prediction_type == :probabilistic &&
	model.is_pure_julia
end

tree = @load DecisionTreeClassifier to load code and instantiate "DecisionTreeClassifier" model

tree2 = DecisionTreeClassifier(max_depth=2) instantiates a model type already in scope

ridge = @load RidgeRegressor pkg=MultivariateStats loads and instantiates a model provided by multiple packages

Scitypes and coercion

scitype(x) is the scientific type of x. For example scitype(2.4) = Continuous

Figure and Table for scalar scitypes

Use schema(X) to get the column scitypes of a table X

coerce(y, Multiclass) attempts coercion of all elements of y into scitype Multiclass

coerce(X, :x1 => Continuous, :x2 => OrderedFactor) to coerce columns :x1 and :x2 of table X.

Ingesting data

Splitting any table into target and input (note semicolon):

using RDatasets
channing = dataset("boot", "channing")
y, X =  unpack(channing,
               ==(:Exit),            # y is the :Exit column
               !=(:Time);            # X is the rest, except :Time
               :Exit=>Continuous,    # correct wrong scitypes
               :Entry=>Continuous,
               :Cens=>Multiclass)

Splitting row indices into train/validation/test:

train, valid, test = partition(eachindex(y), 0.7, 0.2, shuffle=true, rng=1234) for 70:20:10 ratio

Machine construction

Supervised case:

model = KNNRegressor(K=1) and mach = machine(model, X, y)

Unsupervised case:

model = OneHotEncoder() and mach = machine(model, X)

Fitting

fit!(mach, rows=1:100, verbosity=1, force=false)

Prediction

Supervised case: predict(mach, Xnew) or predict(mach, rows=1:100)

Similarly, for probabilistic models: predict_mode, predict_mean and predict_median.

Unsupervised case: transform(mach, rows=1:100) or inverse_transform(mach, rows), etc.

Inspecting objects

@more gets detail on last object in REPL

params(model) gets nested-tuple of all hyperparameters, even nested ones

info(ConstantRegresssor()), info("PCA"), info("RidgeRegressor", pkg="MultivariateStats") gets all properties (aka traits) of registered models

info(rms) gets all properties of a performance measure

schema(X) get column names, types and scitypes, and nrows, of a table X

scitype(model), scitype(rms), scitype(X) gets scientific type of a model, measure or table (encoding key properties)

fitted_params(mach) gets learned parameters of fitted machine

report(mach) gets other training results (e.g. feature rankings)

Resampling strategies

Holdout(fraction_train=…, shuffle=false) for simple holdout

CV(nfolds=6, shuffle=false) for cross-validation

or a list of pairs of row indices:

[(train1, eval1), (train2, eval2), ... (traink, evalk)]

Performance estimation

evaluate(model, X, y, resampling=CV(), measure=rms, operation=predict, weights=..., verbosity=1) evaluate!(mach, resampling=Holdout(), measure=[rms, mav], operation=predict, weights=..., verbosity=1) evaluate!(mach, resampling=[(fold1, fold2), (fold2, fold1)], measure=rms)

Ranges for tuning

If r = range(KNNRegressor(), :K, lower=1, upper = 20, scale=:log) then iterator(r, 6) = [1, 2, 3, 6, 11, 20]

Non-numeric ranges: r = range(model, :parameter, values=…).

Nested ranges: Use dot syntax, as in r = range(EnsembleModel(atom=tree), :(atom.max_depth), ...)

Tuning strategies

Grid(resolution=10) for grid search

Tuning model wrapper

tuned_model = TunedModel(model=…, tuning=Grid(), resampling=Holdout(), measure=…, operation=predict, ranges=…, minimize=true, full_report=true)

Learning curves

curve = learning_curve!(mach, resolution=30, resampling=Holdout(), measure=…, operation=predict, range=…, n=1)

If using Plots.jl:

plot(curve.parameter_values, curve.measurements, xlab=curve.parameter_name, xscale=curve.parameter_scale)

Built-in performance measures

l1, l2, mav, rms, rmsl, rmslp1, rmsp, misclassification_rate, cross_entropy

info(rms) to list properties (aka traits) of the rms measure

using LossFunctions to use more measures

Transformers

Built-ins include: Standardizer, OneHotEncoder, UnivariateBoxCoxTransformer, FeatureSelector, UnivariateStandardizer

Externals include: PCA (in MultivariateStats), KMeans, KMedoids (in Clustering).

Full list: do models(m -> !m[:is_supervised])

Ensemble model wrapper

EnsembleModel(atom=…, weights=Float64[], bagging_fraction=0.8, rng=GLOBAL_RNG, n=100, parallel=true, out_of_bag_measure=[])

Pipelines

With point predictions:

pipe = @pipeline MyPipe(hot=OneHotEncoder(), knn=KNNRegressor(K=3), target=UnivariateStandardizer())

With probabilistic-predictions:

pipe = @pipeline MyPipe(hot=OneHotEncoder(), knn=KNNRegressor(K=3), target=v->log.(V), inverse=v->exp.(v)) is_probabilistic=true

Unsupervised:

pipe = @pipeline MyPipe(stand=Standardizer(), hot=OneHotEncoder())

Define a supervised learning network:

Xs = source(X) ys = source(y, kind=:target)

... define further nodal machines and nodes ...

yhat = predict(knn_machine, W, ys) (final node)

Exporting a learning network as stand-alone model:

Supervised, with final node yhat returning point-predictions:

@from_network Composite(pca=network_pca, knn=network_knn) <= yhat

Supervised, with yhat final node returning probabilistic predictions:

@from_network Composite(knn=network_knn) <= yhat is_probabilistic=true

Unsupervised, with final node Xout:

@from_network Composite(pca=network_pca) <= Xout