def print_property_values(labels, values):
for idx, value in enumerate(values):
print(f"{labels[idx]}: {value}")
print()
# Data rows are in the format of [difficulty_rating],[bpm],[total_length],[cs],[drain],[accuracy],[ar].
labels = [
DIFFICULTY_LABEL, BPM_LABEL, LENGTH_LABEL, CS_LABEL, DRAIN_LABEL,
ACCURACY_LABEL, AR_LABEL
]
filename_labels = []
for label in labels:
filename_labels.append(label.lower().replace(" ", "_"))
# Keep track of each property in separate rows.
points = np.transpose(models_util.load_metadata_dataset())
mins = points.min(axis=-1)
maxes = points.max(axis=-1)
means = np.mean(points, axis=-1)
print("Minimum values:")
print_property_values(labels, mins)
print("Maximum values:")
print_property_values(labels, maxes)
print("Mean values:")
print_property_values(labels, means)
# Plot graphs for each input output feature pair.
from sklearn.neural_network import MLPRegressor
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import StandardScaler
from models import models_util
dataset = models_util.load_metadata_dataset()
# Difficulty and BPM input features.
X_train = dataset[:, 0:2]
y = dataset[:, 4]
model = make_pipeline(StandardScaler(), MLPRegressor(hidden_layer_sizes=(7, )))
model.fit(X_train, y)
models_util.save_model(model, models_util.MetadataPredictor.DRAIN)
print("Trained model saved.")