Created
May 25, 2018 22:46
-
-
Save HeshamMeneisi/a3afea0b7454c8af8b41a4f8e2f29d2a to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
def series_to_supervised(df, n_in=1, n_out=1, targets=[], dropnan=True): | |
""" | |
Converts a time series Pandas DataFrame into a supervised learning problem | |
returns: X(t-n_in+1,...,t), y(t+1,....,t+n_out) | |
""" | |
assert n_in > 0 and n_out > 0 | |
assert all([t in df.columns for t in targets]) | |
n_vars = len(df.columns) | |
cols, names = list(), list() | |
# input sequence (t-n_in+1, ... t-1) | |
for i in range(n_in-1, 0, -1): | |
cols.append(df.shift(i)) | |
names += [(df.columns[j]+'(t-%d)' % (i)) for j in range(n_vars)] | |
# current point in time t | |
cols.append(df) | |
names += [(df.columns[j]+'(t)') for j in range(n_vars)] | |
# forecast sequence (t+1, ... t+n_out) | |
for i in range(1, n_out+1): | |
cols.append(df.shift(-i)[targets]) | |
names += [(c+'(t+%d)' % (i)) for c in targets] | |
# put it all together | |
agg = pd.concat(cols, axis=1) | |
agg.columns = names | |
# drop rows with NaN values | |
if dropnan: | |
agg.dropna(inplace=True) | |
s = len(targets)*n_out | |
return (agg.loc[:, agg.columns[:-s]], agg.loc[:, agg.columns[-s:]]) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment