多列条件值修正：新增df["E"]列填充指定用户时段值至当日其余时段

Got it, let's work through this problem together. You need to add a new column E to your DataFrame where for users Joseph and Emma, we take their 15:00:00 value from column D and propagate that value to all other time slots on the same day for those users. Here's a straightforward way to do this with pandas:

Step 1: Setup and Sample Data #

First, let's assume your DataFrame has columns like User, Datetime, and D (adjust column names if yours differ). Here's a sample to test with:

import pandas as pd

# Sample input DataFrame
data = {
    "User": ["Joseph", "Joseph", "Emma", "Emma", "John", "John"],
    "Datetime": pd.to_datetime([
        "2024-05-20 10:00:00", "2024-05-20 15:00:00",
        "2024-05-20 12:00:00", "2024-05-20 15:00:00",
        "2024-05-20 15:00:00", "2024-05-20 18:00:00"
    ]),
    "D": [10, 25, 15, 30, 20, 22]
}
df = pd.DataFrame(data)

Step 2: Extract Date and Time for Target Filtering #

We first split the Datetime column into date and time components to easily target the 15:00:00 records:

# Extract date and time as separate columns (temporary, for matching)
df["Date"] = df["Datetime"].dt.date
df["Time"] = df["Datetime"].dt.time

Step 3: Create a Mapping of Daily 15:00 Values for Target Users #

Next, we pull out the 15:00 D values for Joseph and Emma, which will be our fill values for their daily records:

# Filter rows where user is Joseph/Emma and time is 15:00:00
target_mask = (df["User"].isin(["Joseph", "Emma"])) & (df["Time"] == pd.to_datetime("15:00:00").time())
daily_fill_values = df[target_mask][["User", "Date", "D"]].rename(columns={"D": "E"})

Step 4: Merge Fill Values Back to Original DataFrame #

We merge this mapping back to the original DataFrame to populate the E column for matching user-date pairs:

# Merge to get the fill values into the E column
df = df.merge(daily_fill_values, on=["User", "Date"], how="left")

# For non-target users (like John), set E to their original D value (adjust if you want NaN instead)
df["E"] = df.apply(lambda row: row["D"] if row["User"] not in ["Joseph", "Emma"] else row["E"], axis=1)

# Clean up temporary columns (optional)
df = df.drop(columns=["Date", "Time"])

Final Result #

Running the code above will give you this output:

User            Datetime   D     E
0  Joseph 2024-05-20 10:00:00  10  25.0
1  Joseph 2024-05-20 15:00:00  25  25.0
2    Emma 2024-05-20 12:00:00  15  30.0
3    Emma 2024-05-20 15:00:00  30  30.0
4    John 2024-05-20 15:00:00  20  20.0
5    John 2024-05-20 18:00:00  22  22.0

Note: If your Datetime column isn't already a datetime type, make sure to convert it first with df["Datetime"] = pd.to_datetime(df["Datetime"]).

内容的提问来源于stack exchange，提问作者Tie_24