将上一列值与当前值相加

Question

I am trying to figure out the logic to write the python code for below requirement. Basically what I want the expected output is, if IDLE != 0 , 'Duration' value should be above cell value + idle. I will be getting a live data with data resolution of 5 minutes. I want to send the duration to a collection. How can I create a logic for this?

Consider below logic (not syntax) I came up until now:

if (df["SPEED"].mean()) != 0:
    df["DURATION"] = 0
    for i in range (len(df)):
        if (df.loc[i, "IDLE"]) == 0:
            df["DURATION"] = df["DISTANCE"]/df["SPEED"]

for i in range (len(df)):
    if (df.loc[i, "DURATION"]) == 0:
        (df.loc[i, "DURATION"]) = ((df.loc[i-1, "DURATION"])+(df.loc[i, "IDLE"]))
.
.

The above code I written is not working. if wondering why if mean(speed)!=0 ? so that initially when speed = 0, I don't want to send output How should I proceed next?

   DISTANCE  SPEED  IDLE  DURATION  EXPECTED_OUTPUT
0         X    0.0     5         0                0
1         X    0.0     5         0                0
2         X    0.2     0        10               10
3         X    0.2     0        10               10
4         X    0.2     0        10               10
5         X    0.2     0        10               10
6         X    0.0     5         0               15
7         X    0.0     5         0               20
8         X    0.2     0        10               10
9         X    0.2     0        10               10
10        X    0.2     0        10               10
11        X    0.2     0        10               10

Answer 1

i worked out a solution

for i in range (1,len(df)):
    if (df.loc[i, "DURATION"]) == 0:
        (df.loc[i, "DURATION"]) = ((df.loc[i-1, "DURATION"])+(df.loc[i, "IDLE"]))

Answer 2

One solution might be to iterate over the dataframe.

1. Find the index of the first speed other than 0. There are a lot of ways to do that, you can try the following:

df[df["SPEED"] != 0].index[0]

2. Copy the Duration column in the EXPECTED_OUTPUT column:

df["EXPECTED_OUTPUT"] = df["DURATION"]

3. Iterate over the dataframe rows starting at the index find in step 1.(here I make the assumption that index rows are sorted).

   1. Check if the duration is 0:

   2. If 1 is verified, verified a previous row exists:

      -> YES: add the previous EXPECTED_OUTPUT cell to current IDLE

      -> NO: set the value you want (here NaN). Set nothing to let the DURATION value

---

Full code :

# Find index where speed starting being != 0
index = df[df["SPEED"] != 0].index[0]
print(index)
# 2

# Copy duration column in expected output
df["EXPECTED_OUTPUT"] = df["DURATION"]

# Iterate over duration row
for i in range(index, len(df)):
    # If duration == 0
    if df.loc[i, 'DURATION'] == 0:
        # Check a previous row exist
        if i > 0:
            df.loc[i, 'EXPECTED_OUTPUT'] = df.loc[i, "IDLE"] + df.loc[i - 1, "EXPECTED_OUTPUT"]
        # If previous row doesn't exist : set NaN
        # You can remove this else statement to have the "duration" value
        # Depends on you
        else:
            df.loc[i, 'EXPECTED_OUTPUT'] = np.NaN

print(df)
#    DISTANCE  SPEED  IDLE  DURATION  EXPECTED_OUTPUT
# 0         X    0.0     5         0                0
# 1         X    0.0     5         0                0
# 2         X    0.2     0        10               10
# 3         X    0.2     0        10               10
# 4         X    0.2     0        10               10
# 5         X    0.2     0        10               10
# 6         X    0.0     5         0               15
# 7         X    0.0     5         0               20
# 8         X    0.2     0        10               10
# 9         X    0.2     0        10               10
# 10        X    0.2     0        10               10
# 11        X    0.2     0        10               10