Hi!
I am trying to submit to the 2sigma news kaggle competition but I cannot figure out any idea how to implement the score-function (https://www.kaggle.com/c/two-sigma-financial-news#evaluation). By now I am running the training using a rmspe score-function:
def inv_y(a): return np.exp(a)
def exp_rmspe(y_pred, targ):
targ = inv_y(targ)
pct_var = (targ - inv_y(y_pred))/targ
return math.sqrt((pct_var**2).mean())
Has anyone an idea to share?
def make_predictions(predictions_template_df, market_obs_df, news_obs_df, is_train, le):
X = get_xy(market_obs_df, news_obs_df, cat_vars, is_train, le)
preds = m.predict(X.values)
predictions_template_df.confidenceValue = m.predict(X)
days = env.get_prediction_days()
for (market_obs_df, news_obs_df, predictions_template_df) in days:
make_predictions(predictions_template_df, market_obs_df, news_obs_df, False, emb_szs)
env.predict(predictions_template_df)
print('Done!')
When I run the predict function on the test data from the kaggle environment I get this error:
