Pivot matrix to time-series – Python
Question:
I’ve got a dataframe with date as first column and time as the name of the other columns.
Date
13:00
14:00
15:00
16:00
…
2022-01-01
B
R
M
M
…
2022-01-02
B
B
B
M
…
2022-01-03
R
B
B
M
…
How could I transform that matrix into a datetime time-series? My objective its something like this:
Date
Data
2022-01-01 13:00
B
2022-01-01 14:00
R
2022-01-01 15:00
M
2022-01-01 16:00
M
…
…
I think it could be done using pivot. I would really appreciate any help you could give me. Thanks in advance!!
Answers:
Try .set_index/.stack. The rest is just convert the string to DateTime:
df = df.set_index("Date").stack().reset_index()
df["Date"] = pd.to_datetime(df["Date"] + " " + df["level_1"])
df = df.rename(columns={0: "Data"})
print(df[["Date", "Data"]])
Prints:
Date Data
0 2022-01-01 13:00:00 B
1 2022-01-01 14:00:00 R
2 2022-01-01 15:00:00 M
3 2022-01-01 16:00:00 M
4 2022-01-02 13:00:00 B
5 2022-01-02 14:00:00 B
6 2022-01-02 15:00:00 B
7 2022-01-02 16:00:00 M
8 2022-01-03 13:00:00 R
9 2022-01-03 14:00:00 B
10 2022-01-03 15:00:00 B
11 2022-01-03 16:00:00 M
An alternative:
df = pd.DataFrame({'Date': ['2022-01-01', '2022-01-02', '2022-01-03'], '13:00': ['B', 'B', 'R'], '14:00': ['R', 'B', 'R'], '15:00': ['M', 'B', 'B'], '16:00': ['M', 'M', 'M']})
df = df.melt(id_vars='Date', var_name='Time', value_name='Data')
df['Date'] = df['Date'] + ' ' + df['Time']
df = df[['Date', 'Data']]
I’ve got a dataframe with date as first column and time as the name of the other columns.
| Date | 13:00 | 14:00 | 15:00 | 16:00 | … |
|---|---|---|---|---|---|
| 2022-01-01 | B | R | M | M | … |
| 2022-01-02 | B | B | B | M | … |
| 2022-01-03 | R | B | B | M | … |
How could I transform that matrix into a datetime time-series? My objective its something like this:
| Date | Data |
|---|---|
| 2022-01-01 13:00 | B |
| 2022-01-01 14:00 | R |
| 2022-01-01 15:00 | M |
| 2022-01-01 16:00 | M |
| … | … |
I think it could be done using pivot. I would really appreciate any help you could give me. Thanks in advance!!
Try .set_index/.stack. The rest is just convert the string to DateTime:
df = df.set_index("Date").stack().reset_index()
df["Date"] = pd.to_datetime(df["Date"] + " " + df["level_1"])
df = df.rename(columns={0: "Data"})
print(df[["Date", "Data"]])
Prints:
Date Data
0 2022-01-01 13:00:00 B
1 2022-01-01 14:00:00 R
2 2022-01-01 15:00:00 M
3 2022-01-01 16:00:00 M
4 2022-01-02 13:00:00 B
5 2022-01-02 14:00:00 B
6 2022-01-02 15:00:00 B
7 2022-01-02 16:00:00 M
8 2022-01-03 13:00:00 R
9 2022-01-03 14:00:00 B
10 2022-01-03 15:00:00 B
11 2022-01-03 16:00:00 M
An alternative:
df = pd.DataFrame({'Date': ['2022-01-01', '2022-01-02', '2022-01-03'], '13:00': ['B', 'B', 'R'], '14:00': ['R', 'B', 'R'], '15:00': ['M', 'B', 'B'], '16:00': ['M', 'M', 'M']})
df = df.melt(id_vars='Date', var_name='Time', value_name='Data')
df['Date'] = df['Date'] + ' ' + df['Time']
df = df[['Date', 'Data']]