Creating an IF statement in Python that looks at previous IF statement output

Question

I am having difficulty creating an IF statement that does the following:

• If C1 = Buy, then Buy
• If C2 = Sell, then Sell
• If C1 & C2 = nan, then the current cell = previous cell

Please see an example below. I am hoping to create a column like 'C3'.

Sample Dataset:

index  C1    C2
0      Buy   nan
1      nan   nan
2      nan   Sell
3      nan   nan
4      Buy   nan
5      nan   Sell
6      nan   Sell
7      nan   nan
8      nan   nan
9      Buy   nan
10     nan   Sell

Output:

index  C1    C2    C3
0      Buy   nan   Buy
1      nan   nan   Buy
2      nan   Sell  Sell
3      nan   nan   Sell
4      Buy   nan   Buy
5      nan   Sell  Sell
6      nan   Sell  Sell
7      nan   nan   Sell
8      nan   nan   Sell
9      Buy   nan   Buy
10     nan   Sell  Sell

Answer 1

You can use pd.DataFrame.ffill along axis=1 followed by pd.Series.ffill:

df['C3'] = df[['C1', 'C2']].ffill(axis=1).iloc[:, -1].ffill()

print(df)

    index   C1    C2    C3
0       0  Buy   NaN   Buy
1       1  NaN   NaN   Buy
2       2  NaN  Sell  Sell
3       3  NaN   NaN  Sell
4       4  Buy   NaN   Buy
5       5  NaN  Sell  Sell
6       6  NaN  Sell  Sell
7       7  NaN   NaN  Sell
8       8  NaN   NaN  Sell
9       9  Buy   NaN   Buy
10     10  NaN  Sell  Sell

Answer 2

Instead of doing the previous if statement, you can simply look at what has been previously put into the c3 list (as that is a result of the previous if statement).

Here is an example of how you can achieve this in python:

c1 = ["Buy", "nan", "nan", "nan", "Buy", "nan", "nan", "nan", "nan", "Buy", "nan"]
c2 = ["nan", "nan", "Sell", "nan", "nan", "Sell", "Sell", "nan", "nan", "nan", "Sell"]

c3 = []
for index in range(len(c1)):
    if c1[index] == "Buy":
        c3.append("Buy")
    elif c2[index] == "Sell":
        c3.append("Sell")
    elif c1[index] == "nan" and c2[index] == "nan": # Implied if reached this point (so else would also suffice here)
        c3.append(c3[index-1]) # look at previous result in list
print(c3)

Output:

['Buy', 'Buy', 'Sell', 'Sell', 'Buy', 'Sell', 'Sell', 'Sell', 'Sell', 'Buy', 'Sell']

Answer 3

Here's a tidy way to do it using Pandas: Swap all the NaN for empty strings, and return whatever string value is in each row. If a row is empty, return what came before it.

import pandas as pd

def decide(data):
    if len(data.sum()):
        return data.sum()
    return decide(df.iloc[data.name - 1])

df.fillna("", inplace=True)
df.apply(decide, axis=1)

Output:

index
0      Buy
1      Buy
2     Sell
3     Sell
4      Buy
5     Sell
6     Sell
7     Sell
8     Sell
9      Buy
10    Sell
dtype: object

Note: Making a couple of assumptions here. First, assuming only Buy or Sell occurs in a row. Second, assuming first row is not empty.

Data:

df = pd.read_clipboard(index_col="index") # copied from OP