I have this data frame:

Node       Band    5-May    12-May    19-May    26-May   2-June
Server1    10000   800      1000      12000     12500    500
Server2    30000   600      3000      12000     12500    500
Server3    17000   1500     12000     12500     500      1000

I need to calculate usage by dividing column values right of “Band” column by the values in “Band” column in pandas.

Resulting data frame should look like this:

df1

Node       Band    5-May    12-May    19-May    26-May   2-June
Server1    10000   0.08     0.10      1.2     1.25       0.05
Server2    30000   0.02     0.1       0.4     0.42       0.02
etc

then I have to count the them if they fall withing certain percentage per node. if any date, usage per node between 70% and 80%, the count for that node is one, however many times the usage is between those thresholds. In this example for Node “Server1”, number of times it breached 100% is 2.

For example:

All summary:

Node    70%-80%  80%-90% 90%-100% 100%+
Server1  0          0       0         2
Server2  0          0       0         0

I can do this by explicitly dividing columns by Band column but the number of columns is not known, it could be many columns.

df['5-May_%']=df['5-May']/df['Band']