So I have a dataframe on Python that has lots of variables happening at once and getting pulled off at the same time and I want to make them into a new dataframe thats grouped by time so for example its like:

but I want to have it as:

I can link the values togther but how do I go about grouping like that over multiple lines?

CodePudding user response：

IIUC, you need to rework your columns, then pivot:

# merge value/unit except for "Machine"
df['value'] = np.where(df['Key'].eq('Machine'),
                       df['Value'],
                       df['Value'].astype(str) ' ' df['Unit'])

# replace "Machine" with "Lid"
df['col'] = df['Key'].replace({'Machine': 'Lid'})

# pivot using the new columns
df2 = (df
 .pivot(index='Date', columns='col', values='value')
 .rename_axis(columns=None).reset_index()
)

output:

                  Date Acceleration  Lid Pressure
0  01/01/2000 12:00:01     10 m/s^2   On  2.0 bar
1  01/01/2000 12:00:02     12 m/s^2  Off  1.5 bar