Seaborn temperature forecasting

TensorFlow
, ,
1

I’m looking at the seaborn temp forecasting. With the modeling, it shifts the time in the past by the future target so you can see the accuracy of the model. I have the model trained and want to pass new data for a prediction but it always forecasts from future target in the past. I have 32 new observations. How do I get it to give me a forecast from now, rather than 6 days ago?

What do I need to do to pass new observations to the model to get current predictions rather than predictions from 6 days ago?

Thanks in advance

I think its coming from multivariate_data:

=====================
def multivariate_data(dataset, target, start_index, end_index, history_size,
target_size, step, single_step=False):
data =
labels = 

  start_index = start_index + history_size
  if end_index is None:
    end_index = len(dataset) - target_size

  for i in range(start_index, end_index):
    indices = range(i-history_size, i, step)
    data.append(dataset[indices])

    if single_step:
      labels.append(target[i+target_size])
    else:
      labels.append(target[i:i+target_size])

  return np.array(data), np.array(labels)

===============================

features_considered = [‘p (mbar)’, ‘T (degC)’, ‘rho (g/m**3)’]
features = df[features_considered]
features.index = df[‘Date Time’]
features.head()

features.plot(subplots=True)

dataset = features.values
dataset

data_mean = dataset[:TRAIN_SPLIT].mean(axis=0)
data_std = dataset[:TRAIN_SPLIT].std(axis=0)

dataset = (dataset-data_mean)/data_std

def multivariate_data(dataset, target, start_index, end_index, history_size,
                      target_size, step, single_step=False):
  data = []
  labels = []

  start_index = start_index + history_size
  if end_index is None:
    end_index = len(dataset) - target_size

  for i in range(start_index, end_index):
    indices = range(i-history_size, i, step)
    data.append(dataset[indices])

    if single_step:
      labels.append(target[i+target_size])
    else:
      labels.append(target[i:i+target_size])

  return np.array(data), np.array(labels)

#past_history = 720
#future_target = 72
#STEP = 6
#past_history = 64
past_history = 32
future_target = 6
STEP = 1
4

Hi Kimia,
I was importing known size data so I just set the start index to be len of my dataset - future_target. I think part of it was the defined variables might have been lost in translation between different users as history_size and target_size are not defined variables. I assume they meant past_history and future_target, but it was never throwing errors.

5

Hi Kimia,
That was not it. I am still off by multiple steps. Is there anyone who can explain how to train the model without going back future_target steps? I was using past history of 32, but it is still predicting from 6 days in the past.