Generate a new column in pandas - conditionaly increase integer count

Question

I would like to have a dataframe of pandas like following: Required Dataframe

Column A - according to these values i would like to generate a new variable - Column B.

All the time, when the values change, it should be "Step_(new_number)" made. So the first zero values are Step_1, then the next values should be Step_2, then other row with numbers should be Step_3 and so on.

This is what i have done so far:

def f(row):
    if row['A'] > 0 :
        val = "Step_1"
    else:
        val = "Step_0"
    return val
df['B'] = df.apply(f, axis=1)

But i do not understand, how to conditionaly increase the count of the values in the new column B. Please note that Step_1 could be replaced by "1" - it does not need to be a text, in order to make the solution easier.

Answer 1

One option is to create some global variables to track the value and the step and reference/update those inside the function that you use with pandas.Series.apply like so:

import pandas as pd

df = pd.DataFrame(dict(
    A = [0,0,0,44.67,44.67,0,0,35.49,35.49,35.49,0]
))

step = 0
value = None

def get_step(x):
    global step
    global value
    if x != value:
        value = x
        step += 1
    return f'Step_{step}'

df['B'] = df['A'].apply(get_step)
print(df)

Output:

        A       B
0    0.00  Step_1
1    0.00  Step_1
2    0.00  Step_1
3   44.67  Step_2
4   44.67  Step_2
5    0.00  Step_3
6    0.00  Step_3
7   35.49  Step_4
8   35.49  Step_4
9   35.49  Step_4
10   0.00  Step_5

Example code in python tutor