Error using prediction with ARIMA (python statsmodels)

Question

I have dataset with dates and random values. The following code is generating the dataset.

def random_dates(start, end, n=10):
        start_u = start.value//10**9
        end_u = end.value//10**9
        return pd.DatetimeIndex((10**9*np.random.randint(start_u, end_u, n)))

    start = pd.to_datetime('2017-01-01')
    end = pd.to_datetime('2018-01-12')

    days = random_dates(start, end)

    np.random.seed(seed=1111)
    data = np.random.randint(1, high=100, size=len(days))
    df = pd.DataFrame({'Date': days, 'values': data})
    df = df.set_index('Date')
    print(df)

I am trying to implement ARIMA timeseries model on the dataset to predict for November, 2018 to December, 2018. From the dataset image it can be seen that I put zeros for the dates where I need to predict the values.

I code below is to do prediction. However, I am getting error at line:2 start of get_prediction is not able to get assigned to the date value of dataframe that I want. TypeError: int() argument must be a string, a bytes-like object or a number, not 'Timestamp'

start = pd.to_datetime("2018-01-10")
pred = results.get_prediction(start, dynamic=False)
pred_ci = pred.conf_int()
ax = tc.plot(label='observed')
pred.predicted_mean.plot(ax=ax, label='One-step ahead Forecast', alpha=.7, figsize=(14, 7))
ax.fill_between(pred_ci.index,
                pred_ci.iloc[:, 0],
                pred_ci.iloc[:, 1], color='k', alpha=.2)
ax.set_xlabel('Date')
ax.set_ylabel('TC')
plt.legend()
plt.show()

I tried to change the date-index to string but it is giving the same error. Kindly tell me where I went wrong. Thanks

cfulton · Accepted Answer · 2018-09-25 22:17:47Z

0

I think this is the result of a bug in Statsmodels that was only recently fixed (https://github.com/statsmodels/statsmodels/pull/5250).

In the meantime, you might try the following (assuming your data is in the variable df)

start = df.index.get_loc(pd.to_datetime("2018-01-10"))

answered Sep 25, 2018 at 22:17

cfulton

3,2052 gold badges17 silver badges16 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Error using prediction with ARIMA (python statsmodels)

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related