What is the variable ypred meaning in the notebook of beat pricing model?

Alphathon
#1

This is in the part of making predictions. My whole team is confused about the variable ypred. (ypred is in the line 7 below)


#suppose you have unseen data

unseen = dl.batch(fromRow=5000000, toRow=5100000)

for data in unseen:

  data = data.iloc[:,:-1].astype(float) # drop the target column to simulate the fact it is unseen data.

  data = engineerData(data, ) #apply data transform

  model = lgb.Booster(model_file=‘lgbmModelUbs.txt’) #load up your model

  ypred = model.predict(data)


def myPredictFunc(newData, dataTransformFunc, model):

  data = dataTransformFunc(data) #apply data transform

  return model.predict(data)


pred = myPredictionFunc(testSet, engineerData, model)

#2

It’s the (scalar) value you’re trying to predict using the first 164 columns as features. Each row is a data point, so if the input has shape (n, 164), ypred should be an n-dimensional vector.