makefile

setup:
	python3 -m venv .venv
	source .venv/bin/activate

install:
	git pull
	pip3 install -r requirements.txt
	pip3 install .

install-dev:
	git pull
	pip3 install -r requirements.txt
	pip3 install -r requirements-dev.txt
	pip3 install -e .
	pre-commit install

train:
	drcomp -m evaluate=False dataset=$(dataset) reducer=AE,CAE,kPCA,LLE,ConvAE,PCA wandb.project=drcomp wandb.group=dataset wandb.name=reducer

evaluate:
	drcomp -m evaluate=True max_evaluation_samples=15000 dataset=$(dataset) reducer=AE,CAE,kPCA,LLE,ConvAE,PCA use_pretrained=True wandb.project=drcomp wandb.group=dataset wandb.name=reducer

train-all:
	make train dataset=MNIST && \
	make train dataset=LfwPeople && \
	make train dataset=SwissRoll && \
	make train dataset=TwinPeaks && \
	make train dataset=FER2013 && \
	make train dataset=OlivettiFaces && \
	make train dataset=ICMR && \
	make train dataset=FashionMNIST

evaluate-all:
	make evaluate dataset=MNIST && \
	make evaluate dataset=LfwPeople && \
	make evaluate dataset=SwissRoll && \
	make evaluate dataset=TwinPeaks && \
	make evaluate dataset=FER2013 && \
	make evaluate dataset=OlivettiFaces && \
	make evaluate dataset=ICMR && \
	make evaluate dataset=FashionMNIST

zip-results:
	zip -r models.zip models && \
	zip -r metrics.zip metrics


zip-model-for:
	zip -r $(dataset)-models.zip models/$(dataset)