I have this data frame:
name event join_date created_at
A X 2020-12-01 2020-12-01
A X 2020-12-01 2020-12-01
A X 2020-12-01 2020-12-02
A Y 2020-12-01 2020-12-02
B X 2020-12-05 2020-12-05
B X 2020-12-05 2020-12-07
C X 2020-12-07 2020-12-08
C X 2020-12-07 2020-12-09
...
I want to transform it into this data frame:
name event join_date day_0 day_1 day_2 .... day_n
A X 2020-12-01 2 1 0 0
A Y 2020-12-01 0 1 0 0
B X 2020-12-05 1 0 1 0
C X 2020-12-07 0 1 1 0
...
the first rows mean that user A doing twice Event X on day_0 (first day he joins) and once on the first day and so on until day_n
For now, the result is like this:
name event join_date day_0 day_1 day_2 .... day_n
A X 2020-12-01 2 1 0 0
A Y 2020-12-01 0 1 0 0
B X 2020-12-05 1 0 1 0
C X 2020-12-07 1 1 0 0
...
the code set the 2020-12-02 as day_0, not day_1 because there is no 2020-12-01 on A user with Y event
First subtract all values created_at by first value per groups by GroupBy.transform.
Then use DataFrame.pivot_table first, add all possible datetimes by DataFrame.reindex by timedelta_range and then convert columns names by range:
df['d'] = df['created_at'].sub(df['join_date'])
print (df)
name event join_date created_at d
0 A X 2020-12-01 2020-12-01 0 days
1 A X 2020-12-01 2020-12-01 0 days
2 A X 2020-12-01 2020-12-02 1 days
3 A Y 2020-12-01 2020-12-02 1 days
4 B X 2020-12-05 2020-12-05 0 days
5 B X 2020-12-05 2020-12-07 2 days
6 C X 2020-12-07 2020-12-08 1 days
7 C X 2020-12-07 2020-12-09 2 days
df1 = (df.pivot_table(index=['name','event','join_date'],
columns='d',
aggfunc='size',
fill_value=0)
.reindex(pd.timedelta_range(df['d'].min(), df['d'].max()),
axis=1,
fill_value=0))
df1.columns = [f'day_{i}' for i in range(len(df1.columns))]
df1 = df1.reset_index()
print (df1)
name event join_date day_0 day_1 day_2
0 A X 2020-12-01 2 1 0
1 A Y 2020-12-01 0 1 0
2 B X 2020-12-05 1 0 1
3 C X 2020-12-07 0 1 1
Related
I Have a dataframe as follows:
df = pd.DataFrame({'Key':[1,1,1,1,2,2,2,4,4,4,5,5],
'Activity':['A','A','H','B','B','H','H','A','C','H','H','B'],
'Date':['2022-12-03','2022-12-04','2022-12-06','2022-12-08','2022-12-03','2022-12-06','2022-12-10','2022-12-03','2022-12-04','2022-12-07','2022-12-03','2022-12-13']})
I need to count the activities for each 'Key' that occur before 'Activity' == 'H' as follows:
Required Output
My Approach
Sort df by Key & Date ( Sample input is already sorted)
drop the rows that occur after 'H' Activity in each group as follows:
Groupby df.groupby(['Key', 'Activity']).count()
Is there a better approach , if not then help me in code for dropping the rows that occur after 'H' Activity in each group.
Thanks in advance !
You can bring the H dates "back" into each previous row to use in a comparison.
First mark each H date in a new column:
df.loc[df["Activity"] == "H" , "End"] = df["Date"]
Key Activity Date End
0 1 A 2022-12-03 NaT
1 1 A 2022-12-04 NaT
2 1 H 2022-12-06 2022-12-06
3 1 B 2022-12-08 NaT
4 2 B 2022-12-03 NaT
5 2 H 2022-12-06 2022-12-06
6 2 H 2022-12-10 2022-12-10
7 4 A 2022-12-03 NaT
8 4 C 2022-12-04 NaT
9 4 H 2022-12-07 2022-12-07
10 5 H 2022-12-03 2022-12-03
11 5 B 2022-12-13 NaT
Backward fill the new column for each group:
df["End"] = df.groupby("Key")["End"].bfill()
Key Activity Date End
0 1 A 2022-12-03 2022-12-06
1 1 A 2022-12-04 2022-12-06
2 1 H 2022-12-06 2022-12-06
3 1 B 2022-12-08 NaT
4 2 B 2022-12-03 2022-12-06
5 2 H 2022-12-06 2022-12-06
6 2 H 2022-12-10 2022-12-10
7 4 A 2022-12-03 2022-12-07
8 4 C 2022-12-04 2022-12-07
9 4 H 2022-12-07 2022-12-07
10 5 H 2022-12-03 2022-12-03
11 5 B 2022-12-13 NaT
You can then select rows with Date before End
df.loc[df["Date"] < df["End"]]
Key Activity Date End
0 1 A 2022-12-03 2022-12-06
1 1 A 2022-12-04 2022-12-06
4 2 B 2022-12-03 2022-12-06
7 4 A 2022-12-03 2022-12-07
8 4 C 2022-12-04 2022-12-07
To generate the final form - you can use .pivot_table()
(df.loc[df["Date"] < df["End"]]
.pivot_table(index="Key", columns="Activity", values="Date", aggfunc="count")
.reindex(df["Key"].unique()) # Add in keys with no match e.g. `5`
.fillna(0)
.astype(int))
Activity A B C
Key
1 2 0 0
2 0 1 0
4 1 0 1
5 0 0 0
Try this:
(df.loc[df['Activity'].eq('H').groupby(df['Key']).cumsum().eq(0)]
.set_index('Key')['Activity']
.str.get_dummies()
.groupby(level=0).sum()
.reindex(df['Key'].unique(),fill_value=0)
.reset_index())
Output:
Key A B C
0 1 2 0 0
1 2 0 1 0
2 4 1 0 1
3 5 0 0 0
You can try:
# sort by Key and Date
df.sort_values(['Key', 'Date'], inplace=True)
# this is to keep Key in the result when no values are kept after the filter
df.Key = df.Key.astype('category')
# filter all rows after the 1st H for each Key and then pivot
df[~df.Activity.eq('H').groupby(df.Key).cummax()].pivot_table(
index='Key', columns='Activity', aggfunc='size'
).reset_index()
#Activity Key A B C
#0 1 2 0 0
#1 2 0 1 0
#2 4 1 0 1
#3 5 0 0 0
I have a problem. I want to calculate from a date for example 2022-06-01 how many touches the customer with the customerId == 1 had in the last 6 months. He had two touches 2022-05-25 and 2022-05-20. I have now calculated the date up to which the data should be taken into account. However, I don't know how to group the customer and say the date you have is up to count_from_date how many touches the customer has had.
Dataframe
customerId fromDate
0 1 2022-06-01
1 1 2022-05-25
2 1 2022-05-25
3 1 2022-05-20
4 1 2021-09-05
5 2 2022-06-02
6 3 2021-03-01
7 3 2021-02-01
import pandas as pd
d = {'customerId': [1, 1, 1, 1, 1, 2, 3, 3],
'fromDate': ["2022-06-01", "2022-05-25", "2022-05-25", "2022-05-20", "2021-09-05",
"2022-06-02", "2021-03-01", "2021-02-01"]
}
df = pd.DataFrame(data=d)
print(df)
from datetime import date
from dateutil.relativedelta import relativedelta
def find_last_date(date):
six_months = date + relativedelta(months=-6)
return six_months
df['fromDate'] = pd.to_datetime(df['fromDate'], errors='coerce')
df['count_from_date'] = df['fromDate'].apply(lambda x: find_last_date(x))
print(df)
What I have
customerId fromDate count_from_date
0 1 2022-06-01 2021-12-01
1 1 2022-05-25 2021-11-25
2 1 2022-05-25 2021-11-25
3 1 2022-05-20 2021-11-20
4 1 2021-09-05 2021-03-05
5 2 2022-06-02 2021-12-02
6 3 2021-03-01 2020-09-01
7 3 2021-02-01 2020-08-01
What I want
customerId fromDate count_from_date occur_last_6_months
0 1 2022-06-01 2021-12-01 3 # 2022-05-25, 2022-05-20, 2022-05-20 = 3
1 1 2022-05-25 2021-11-25 1 # 2022-05-20 = 1
2 1 2022-05-25 2021-11-25 1 # 2022-05-20 = 1
3 1 2022-05-20 2021-11-20 0 # No in the last 6 months
4 1 2021-09-05 2021-03-05 0 # No in the last 6 months
5 2 2022-06-02 2021-12-02 0 # No in the last 6 months
6 3 2021-03-01 2020-09-01 1 # 2021-02-01 = 1
7 3 2021-02-01 2020-08-01 0 # No in the last 6 months
You can try groupby customerId and loop through the rows in subgroup to count number of fromDate between fromDate and count_from_date
def count(g):
m = pd.concat([g['fromDate'].between(d1, d2, 'neither')
for d1, d2 in zip(g['count_from_date'], g['fromDate'])], axis=1)
g = g.assign(occur_last_6_months=m.sum().tolist())
return g
out = df.groupby('customerId').apply(count)
print(out)
customerId fromDate count_from_date occur_last_6_months
0 1 2022-06-01 2021-12-01 3
1 1 2022-05-25 2021-11-25 1
2 1 2022-05-25 2021-11-25 1
3 1 2022-05-20 2021-11-20 0
4 1 2021-09-05 2021-03-05 0
5 2 2022-06-02 2021-12-02 0
6 3 2021-03-01 2020-09-01 1
7 3 2021-02-01 2020-08-01 0
For this problem, the challenge for a performant solution is to manipulate the data as to have an appropriate structure to run rolling window operations on it.
First of all, we need to avoid having duplicate indices. In your case, this means aggregating multiple touch points in a single day:
>>> df = df.groupby(['customerId', 'fromDate'], as_index=False).count()
customerId fromDate count_from_date
0 1 2021-09-05 1
1 1 2022-05-20 1
2 1 2022-05-25 2
3 1 2022-06-01 1
4 2 2022-06-02 1
5 3 2021-02-01 1
6 3 2021-03-01 1
Now, we can set the index to fromDate, sort it and groupby customerId as to be able to use rolling windows. I here use a 180D rolling window (6 months):
>>> roll_df = df.set_index(['fromDate'])
.sort_index()
.groupby('customerId')
.apply(lambda s: s['count_from_date'].rolling('180D').sum())
The sort_index step is important to ensure your data is monotonically increasing.
However, this also counts the touch on the day itself, which seems not what you want, so we remove 1 from the result:
>>> roll_df - 1
customerId fromDate
1 2021-09-05 0.0
2022-05-20 0.0
2022-05-25 2.0
2022-06-01 3.0
2 2022-06-02 0.0
3 2021-02-01 0.0
2021-03-01 1.0
Name: count_from_date, dtype: float64
Finally, we divide by the initial counts to get back to the original structure:
>>> roll_df / df.set_index(['customerId', 'fromDate'])['count_from_date']
customerId fromDate count_from_date
0 1 2021-09-05 0.0
1 1 2022-05-20 0.0
2 1 2022-05-25 1.0
3 1 2022-06-01 3.0
4 2 2022-06-02 0.0
5 3 2021-02-01 0.0
6 3 2021-03-01 1.0
You can always .reset_index() at the end.
The one liner solution is
(df.set_index(['fromDate'])
.sort_index()
.groupby('customerId')
.apply(lambda s: s['count_from_date'].rolling('180D').sum())
- 1) / df.set_index(['customerId', 'fromDate'])['count_from_date']
I have a list of products produced or processes finished like this one:
Name
Timestamp Start
Timestamp Stop
Product 1
2021-01-01 15:15:00
2021-01-01 15:37:00
Product 1
2021-01-01 15:30:00
2021-01-01 15:55:00
Product 1
2021-01-02 15:05:00
2021-01-02 15:22:00
Product 1
2021-01-03 15:45:00
2021-01-03 15:55:00
...
...
...
What I want to do is to calculate the amount of time where no product/process happened in a given timeframe, for example from 15:00 to 16:00 and, to be more specific, each day.
The output could be "amount of idle minutes/time where nothing happened" or "percentage of idle time".
import pandas as pd
import datetime
df = pd.read_csv('example_data.csv')
# generate list of products
listOfProducts = df['NAME'].drop_duplicates().tolist()
# define timeframe for each day
startTime = datetime.time(15, 0)
stopTime = datetime.time(16, 0)
# define daterange to look for
startDay = datetime(2021, 1, 1)
stopDay = datetime(2021,1, 5)
# do it for every product
for i in listOfProducts:
# filter dataframe by product
df_product = df[df['NAME'] == i]
# sort dataframe by start
df_product = df_product.sort_values(by='started')
# ... how to proceed?
The wanted output should look like this or similiar:
Day
Time idle
2021-01-01
00:20:00
2021-02-01
00:43:00
2021-03-01
00:50:00
...
...
Here are some notes that are important:
Timeranges of products can overlap between each other, in this case they should only "count once"
Timeranges of products can overlap the borders (15:00 or 16:00 in this case), in this case the time within the borders should be counted
I struggle to implement it in a pandas-way, because this border-cases prevent me from adding up Timedeltas.
In the past, I solved this issue by iterating row by row from here and adding the minutes or seconds. But I'm sure there is a more pandas-way, maybe with the .groupby() function?
Input data:
>>> df
Name Start Stop
0 Product 1 2021-01-01 14:49:00 2021-01-01 15:04:00 # OK (overlap 4')
1 Product 1 2021-01-01 15:15:00 2021-01-01 15:37:00 # OK
2 Product 1 2021-01-01 15:30:00 2021-01-01 15:55:00 # OK
3 Product 1 2021-01-02 15:05:00 2021-01-02 15:22:00 # OK
4 Product 1 2021-01-03 15:45:00 2021-01-03 15:55:00 # OK
5 Product 1 2021-01-03 15:51:00 2021-01-03 16:23:00 # OK (overlap 9')
6 Product 1 2021-01-04 14:28:00 2021-01-04 17:12:00 # OK (overlap 60')
7 Product 1 2021-01-05 11:46:00 2021-01-05 13:40:00 # Out of bounds
8 Product 1 2021-01-05 17:20:00 2021-01-05 19:11:00 # Out of bounds
First, remove data out of bounds (7 & 8):
import datetime
START = datetime.time(15)
STOP = datetime.time(16)
df1 = df.loc[(df["Start"].dt.floor(freq="H").dt.time <= START)
& (START <= df["Stop"].dt.floor(freq="H").dt.time),
["Start", "Stop"]]
Extract the minute of Start and Stop datetime. If the process began before 15:00, set to 0 because we want only keep overlap part. If the process ended after 16:00, set the minute to 59.
import numpy as np
df1["m1"] = np.where(df1["Start"].dt.time > START,
df1["Start"].sub(df1["Start"].dt.floor(freq="H"))
.dt.seconds // 60, 0)
df1["m2"] = np.where(df1["Stop"].dt.time < STOP,
df1["Stop"].sub(df1["Stop"].dt.floor(freq="H"))
.dt.seconds // 60, 59)
>>> df1
Start Stop m1 m2
0 2021-01-01 14:49:00 2021-01-01 15:04:00 0 4
1 2021-01-01 15:15:00 2021-01-01 15:37:00 15 37
2 2021-01-01 15:30:00 2021-01-01 15:55:00 30 55
3 2021-01-02 15:05:00 2021-01-02 15:22:00 5 22
4 2021-01-03 15:45:00 2021-01-03 15:55:00 45 55
5 2021-01-03 15:51:00 2021-01-03 16:23:00 51 59
6 2021-01-04 14:28:00 2021-01-04 17:12:00 0 59
Create an empty table len(df1)x60' to store process usage:
out = pd.DataFrame(0, index=df1.index, columns=pd.RangeIndex(60))
Fill the out dataframe:
for idx, (i1, i2) in df1[["m1", "m2"]].iterrows():
out.loc[idx, i1:i2] = 1
>>> out
0 1 2 3 4 5 6 ... 53 54 55 56 57 58 59
0 1 1 1 1 1 0 0 ... 0 0 0 0 0 0 0 # 4'
1 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 ... 1 1 1 0 0 0 0
3 0 0 0 0 0 1 1 ... 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 ... 1 1 1 0 0 0 0
5 0 0 0 0 0 0 0 ... 1 1 1 1 1 1 1
6 1 1 1 1 1 1 1 ... 1 1 1 1 1 1 1 # full hour
[7 rows x 60 columns]
Finally, compute the idle minutes:
>>> 60 - (out.groupby(df1["Start"].dt.date).sum() & 1).sum(axis="columns")
Start
2021-01-01 22
2021-01-02 42
2021-01-03 50
2021-01-04 0
dtype: int64
Note: you have to determine if the Stop datetime is closed or not.
I want to group nearby dates together, using a rolling window (?) of three week periods.
See example and attempt below:
import pandas as pd
d = {'id':[1, 1, 1, 1, 2, 3],
'datefield':['2021-01-01', '2021-01-15', '2021-01-30', '2021-02-05', '2020-02-10', '2020-02-20']}
df = pd.DataFrame(data=d)
df['datefield'] = pd.to_datetime(df['datefield'])
# id datefield
#0 1 2021-01-01
#1 1 2021-01-15
#2 1 2021-02-01
#3 2 2020-02-10
#4 3 2020-02-20
df['event'] = df.groupby(['id', pd.Grouper(key='datefield', freq='3W')]).ngroup()
# id datefield event
#0 1 2021-01-01 0
#1 1 2021-01-15 0
#2 1 2021-01-30 1 #Should be 0, since last id 1 event happened just 2 weeks ago
#3 1 2021-02-05 1 #Should be 0
#4 2 2020-02-10 2
#5 3 2020-02-20 3 #Correct, within 3 weeks of another but since the ids are not the same the event is different
Can compute different columns to make it easily understandable
df
id datefield
0 1 2021-01-01
1 1 2021-01-15
2 1 2021-01-30
3 1 2021-02-05
4 2 2020-02-10
5 2 2020-03-20
Calculate difference between dates in number of days
df['diff'] = df['datefield'].diff().dt.days
Get previous ID
df['prevId'] = df['id'].shift()
Decide whether to increment or not
df['increment'] = np.where((df['diff']>21) | (df['prevId'] != df['id']), 1, 0)
Lastly, just get the cumulative sum
df['event'] = df['increment'].cumsum()
Output
id datefield diff prevId increment event
0 1 2021-01-01 NaN NaN 1 1
1 1 2021-01-15 14.0 1.0 0 1
2 1 2021-01-30 15.0 1.0 0 1
3 1 2021-02-05 6.0 1.0 0 1
4 2 2020-02-10 -361.0 1.0 1 2
5 2 2020-03-20 39.0 2.0 1 3
Let's try a different approach using a boolean series instead:
df['group'] = ((df['datefield'].diff()
.fillna(pd.Timedelta(1))
.gt(pd.Timedelta(weeks=3))) |
(df['id'].ne(df['id'].shift()))).cumsum()
Output:
id datefield group
0 1 2021-01-01 1
1 1 2021-01-15 1
2 1 2021-01-30 1
3 1 2021-02-05 1
4 2 2020-02-10 2
5 2 2020-03-20 3
Is the difference between the previous row greater than 3 weeks:
print((df['datefield'].diff()
.fillna(pd.Timedelta(1))
.gt(pd.Timedelta(weeks=3))))
0 False
1 False
2 False
3 False
4 False
5 True
Name: datefield, dtype: bool
Or is the current id not equal to the previous id:
print((df['id'].ne(df['id'].shift())))
0 True
1 False
2 False
3 False
4 True
5 False
Name: id, dtype: bool
or (|) together the conditions
print((df['datefield'].diff()
.fillna(pd.Timedelta(1))
.gt(pd.Timedelta(weeks=3))) |
(df['id'].ne(df['id'].shift())))
0 True
1 False
2 False
3 False
4 True
5 True
dtype: bool
Then use cumsum to increment every where there is a True value to delimit the groups.
*Assumes id and datafield columns are appropriately ordered.
It looks like you want the diff between consecutive rows to be three weeks or less, otherwise a new group is formed. You can do it like this, starting from initial time t0:
df = df.sort_values("datefield").reset_index(drop=True)
t0 = df.datefield.iloc[0]
df["delta_t"] = pd.TimedeltaIndex(df.datefield - t0)
df["group"] = (df.delta_t.dt.days.diff() > 21).cumsum()
output:
id datefield delta_t group
0 2 2020-02-10 0 days 0
1 2 2020-03-20 39 days 1
2 1 2021-01-01 326 days 2
3 1 2021-01-15 340 days 2
4 1 2021-01-30 355 days 2
5 1 2021-02-05 361 days 2
Note that your original dataframe is not sorted properly.
I have a dataframe like this
enter image description here
I want to backfill each item where date_activity is 1/1/2000 12:00:00 with the max date_activity for each item_id. In the end, I want something like this using pandas
enter image description here
Create missing values by Series.duplicated and Series.mask and then backfilling values:
df = pd.DataFrame({'item_id':[1,1,1,2,2,2,2],
'date_active':pd.date_range('2019-02-02', periods=7)})
print (df)
item_id date_active
0 1 2019-02-02
1 1 2019-02-03
2 1 2019-02-04
3 2 2019-02-05
4 2 2019-02-06
5 2 2019-02-07
6 2 2019-02-08
df['date_active'] = df['date_active'].mask(df['item_id'].duplicated(keep='last')).bfill()
print (df)
item_id date_active
0 1 2019-02-04
1 1 2019-02-04
2 1 2019-02-04
3 2 2019-02-08
4 2 2019-02-08
5 2 2019-02-08
6 2 2019-02-08
Details:
print (df['item_id'].duplicated(keep='last'))
0 True
1 True
2 False
3 True
4 True
5 True
6 False
Name: item_id, dtype: bool
print (df['date_active'].mask(df['item_id'].duplicated(keep='last')))
0 NaT
1 NaT
2 2019-02-04
3 NaT
4 NaT
5 NaT
6 2019-02-08
Name: date_active, dtype: datetime64[ns]
EDIT:
If real data is necessary sorting values before solution for last maximum value per group:
print (df)
item_id date_active
0 1 7/26/2019 17:06
1 1 8/27/2019 17:06
df['date_active'] = pd.to_datetime(df['date_active'])
df = df.sort_values(['item_id','date_active'])
df['date_active'] = df['date_active'].mask(df['item_id'].duplicated(keep='last')).bfill()
print (df)
item_id date_active
0 1 2019-08-27 17:06:00
1 1 2019-08-27 17:06:00
EDIT1: Use DataFrame.resample for add missing datetimes per groups:
df['date_active'] = pd.to_datetime(df['date_active'])
df = df.sort_values(['item_id','date_active'])
df = (df.set_index('date_active').groupby('item_id')
.resample('D')
.last()
.drop('item_id', axis=1)
.reset_index())
df['date_active'] = df['date_active'].mask(df['item_id'].duplicated(keep='last')).bfill()
print (df.tail())
item_id date_active
28 1 2019-08-27
29 1 2019-08-27
30 1 2019-08-27
31 1 2019-08-27
32 1 2019-08-27