Conditional Cumsum with Reset

cumsum with a condition to restart in R

You may use cumsum to create groups as well.

library(dplyr)

df <- df %>%
  group_by(group = cumsum(dplyr::lag(port == 0, default = 0))) %>%
  mutate(cumsum_G = cumsum(G)) %>%
  ungroup

df

#  inv   ass    port     G group cumsum_G
#  <chr> <chr> <int> <int> <dbl>    <int>
#1 i     x         2     1     0        1
#2 i     x         2     0     0        1
#3 i     x         0     1     0        2
#4 i     x         3     0     1        0
#5 i     x         3     1     1        1

You may remove the group column from output using %>% select(-group).

data

df <- structure(list(inv = c("i", "i", "i", "i", "i"), ass = c("x", 
"x", "x", "x", "x"), port = c(2L, 2L, 0L, 3L, 3L), G = c(1L, 
0L, 1L, 0L, 1L)), class = "data.frame", row.names = c(NA, -5L))

Pandas Cumsum conditional reset

df = pd.DataFrame({'Size':[8,8,8,8,7,6,7,6,5,2]})

ls = []  
cumsum = 0
last_reset = 0
for _, row in df.iterrows():
    if cumsum + row.Size <= 16:
        cumsum += row.Size
    else:
        last_reset = cumsum
        cumsum = row.Size
    ls.append(cumsum)

df['cumsum'] = ls

Result:

    Size    cumsum
0   8       8
1   8       16
2   8       8
3   8       16
4   7       7
5   6       13
6   7       7
7   6       13
8   5       5
9   2       7

Conditional cumsum and reset to 0

Create custom groups:

grps = df['comment'].str.contains(r'^manual input').cumsum()

df['cumulation'] = df.groupby(grps)['pre_cont_diff'].cumsum()

Output:

>>> df
           comment  count  pre_cont_diff  cumulation
0           auto 1     10            0.0         0.0
1           auto 2     30           20.0        20.0
2           auto 3     70           40.0        60.0
3           auto 4    120           50.0       110.0
4           auto 5    120            0.0       110.0
5           auto 6    130           10.0       120.0
6           auto 7    150           20.0       140.0
7   manual input 1    150            0.0         0.0
8           auto 8    200           50.0        50.0
9           auto 9    230           30.0        80.0
10  manual input 2    230            0.0         0.0

Details:

>>> pd.concat([df['comment'], grps], axis=1)
           comment  comment
0           auto 1        0
1           auto 2        0
2           auto 3        0
3           auto 4        0
4           auto 5        0
5           auto 6        0
6           auto 7        0
7   manual input 1        1
8           auto 8        1
9           auto 9        1
10  manual input 2        2

reset cumulative sum based on another column

Here is an option, you first create the tp_cum column and then cumsum()

import pandas as pd
import numpy as np

df = pd.DataFrame([["y",10 ], 
["y",20  ],
["y",5  ],
["n",30   ],
["n",20   ],
["n",5 ],
["y",10  ], 
["y",40  ],
["y",15 ]],columns = ["type","sale"])

df["type2"] = np.cumsum((df["type"] != df["type"].shift(1)))
df["cum_sale"] = df[["sale","type2"]].groupby("type2").cumsum()
df

Output:

    type    sale    type2  cum_sale
0   y       10      1      10
1   y       20      1      30
2   y       5       1      35
3   n       30      2      30
4   n       20      2      50
5   n       5       2      55
6   y       10      3      10
7   y       40      3      50
8   y       15      3      65

Cumulative sum that resets when the condition is no longer met

Note: Uses global variable

c = 0
def fun(x):
    global c    
    if x['speed'] > 2.0:
        c = 0
    else:
        c = x['timedelta']+c
    return c

df = pd.DataFrame( {'datetime': ['1-1-2019 19:30:00']*7,
    'speed': [0.5,.7,0.1,5.0,25.0,0.1,0.1], 'timedelta': [0,2,2,2,2,4,7]})

df['cum_sum']=df.apply(fun, axis=1)

            datetime    speed   timedelta   cum_sum
0   1-1-2019 19:30:00   0.5     0           0
1   1-1-2019 19:30:00   0.7     2           2
2   1-1-2019 19:30:00   0.1     2           4
3   1-1-2019 19:30:00   5.0     2           0
4   1-1-2019 19:30:00   25.0    2           0
5   1-1-2019 19:30:00   0.1     4           4
6   1-1-2019 19:30:00   0.1     7           11

Cumsum Reset based on a condition in Pandas

Use groupby.cumsum:

df['Cumulative'] = df.groupby('TransactionId')['Delta'].cumsum()

print (df)

  TransactionId  Delta  Cumulative
0       14         2       2
1       14         3       5
2       14         1       6
3       14         2       8
4       15         4       4 
5       15         2       6 
6       15         3       9

Conditional cumsum with reset when accumulating and substracting at once

Here is one way you might do this:

ve$DO <- Reduce(function(x,y) pmax(x + y, 0), with(ve, V2-V3*(V1 >  0)), accumulate = TRUE)

ve
   V1 V2 V3 DO
1  -2  6  0  6
2   4  0  5  1
3   3  0  8  0
4  -5  1  0  1
5  -5  0  0  1
6  -7  2  0  3
7  -8  3  0  6
8  -2  5  0 11
9  -3  7  0 18
10 -5  7  0 25
11 -7  8  0 33
12 -8  2  0 35
13 -9  0  0 35
14 -2  0  7 35
15  1  0  9 26
16  2  0 12 14
17  4  0  0 14

Equivalent using purrr/dplyr:

library(purrr)
library(dplyr)

ve %>%
  mutate(DO = accumulate(V2-V3*(V1 >  0), .f = ~pmax(.x + .y, 0)))

---

Related Topics