How to iterate backwards through dataframe until it meets specified condition, then continue iterati-CodePudding

I have a pandas dataframe column called "elements". It either has numbers of length 9 or of length 7 in it's value. The first row of dataframe is always the one with 9 numbers and the last row is always with 7.

Example:

{element:[120000000, 8000000, 120003000, 7000000, 120003333, 120003444, 5000000]

I want to create another column called "assigned". That column has to assign values, going from backwards of the dataframe, taking the 7 number value from column "element" until it meets 9 number value. Then use the 7 number value it used before for the last time and go to a new row to start using next 7 number value.

Here is an example:

{element:[120000000, 8000000, 120003000, 7000000, 120003333, 120003444, 5000000]}
{assigned:[8000000, 8000000, 7000000, 7000000, 5000000, 5000000, 5000000]}

Honestly I am stuck and have no idea how to do it.

CodePudding user response：

Use Series.where for replace values greater like 100000000 to missing values and then back filling next values, last convert to integers:

df['assigned'] = df['element'].where(df['element'].lt(100000000)).bfill().astype(int)
print (df)
     element  assigned
0  120000000   8000000
1    8000000   8000000
2  120003000   7000000
3    7000000   7000000
4  120003333   5000000
5  120003444   5000000
6    5000000   5000000

CodePudding user response：

One option, is to use str.len with bfill :

#is it a 7-digit number ?
m = df["element"].astype(str).str.len().eq(7)

df["assigned"] = df["element"].where(m).bfill(downcast="infer")

Output:

print(df)

     element  assigned
0  120000000   8000000
1    8000000   8000000
2  120003000   7000000
3    7000000   7000000
4  120003333   5000000
5  120003444   5000000
6    5000000   5000000