# Count the max number of consecutive 1 and 0 in Pandas Dataframe [closed]

Hey I have the following Dataset

import pandas as pd
df = pd.DataFrame({
'column1': [0,0,1,0,1,0,0,1,1,0,1,1,1]})


I want to be able to count the number of consecutive 1 and 0 and generate 2 columns as such:

consec0: 1,2,_,1,_,1,2,_,_,1,_,_,_
consec1: _,_,1,_,1,_,_,1,2_,1,2,3


I then want to take the max number of consecutive and create two lists:

max_consec0: 2,1,2,1
max_consec1: 1,1,2,3


My dataset in the end will be just max_consec0 and max_consec1

To check if a value has changed, you can use .diff and check if it's non-zero with .ne(0) (the NaN in the top will be considered different than zero), and then count the changes with .cumsum, like this:

df['counter'] = df.diff().ne(0).cumsum()


Afterward, you can create a second dataframe, where the indices are the groups of consecutive values, and the column values are the value (0 or 1, in your case) and length (which is what you ultimately want):

df2 = df.groupby('counter')['column1'].min().to_frame(name='value').join(
df.groupby('counter')['column1'].count().rename('number'))


The resulting max_consec0, max_consec1 are just the values in the [number] column, filtered by the [value] column:

max_consec0 = df2[df2['value']==0]['number'].tolist()
max_consec1 = df2[df2['value']==1]['number'].tolist()


You can verify that the result is [2, 1, 2, 1] and [1, 1, 2, 3], as desired.

• awesome solution! Jul 13 '20 at 13:03
• Thank you! This is exactly what I was looking for Jul 13 '20 at 14:42

For this sort of problem you can use np.where and multiple boolean expressions to get your answer.

#1 We need to test if the column is equal to your target values, 0 or 1

second, we have to ensure the column value is not equal to the column value above.

3rd, for any value that is not equal to our input value we return a nan as its easier to use with numeric values.

import numpy as np

df['col2'] = np.where(
df["column1"].eq(0),
df.groupby(df.column1.ne(df.column1.shift()).cumsum()).cumcount() + 1,
np.nan,
)

df['col3'] = np.where(
df["column1"].eq(1),
df.groupby(df.column1.ne(df.column1.shift()).cumsum()).cumcount() + 1,
np.nan,
)


print(df)

column1  col2  col3
0         0   1.0   NaN
1         0   2.0   NaN
2         1   NaN   1.0
3         0   1.0   NaN
4         1   NaN   1.0
5         0   1.0   NaN
6         0   2.0   NaN
7         1   NaN   1.0
8         1   NaN   2.0
9         0   1.0   NaN
10        1   NaN   1.0
11        1   NaN   2.0
12        1   NaN   3.0


we need to create a surrogate key for each group of numbers before a blank row and take the max of each group.

df.assign(
key1=df.groupby(df["col2"].isnull())["col2"].transform("cumcount").cumsum()
).groupby("key1")["col2"].max().dropna()

[2.0, 1.0, 2.0, 1.0]

df.assign(
key2=df.groupby(df["col3"].isnull())["col3"].transform("cumcount").cumsum()
).groupby("key2")["col3"].max().dropna().tolist()

[1.0, 1.0, 2.0, 3.0]

• Thank you for taking the time to help me ! Jul 13 '20 at 14:42
• I like the usage of shift here Jul 13 '20 at 17:57

You can try this implementation:

num0=0
num1=0
consec0=[]
consec1=[]
for i in range(len(df)):
if(df.iloc[i,0])==0:
num0=num0+1;
num1=0;
if(df.iloc[i,0])==1:
num0=0;
num1=num1+1;
consec0.append(num0)
consec1.append(num1)
df['consec0']=consec0
df['consec1']=consec1
max_consec0=df['consec0'].max()
max_consec1=df['consec1'].max()

• Thanks this is almost what I need, I want the max consec columns to be split by either 0 or 1. So that if I have 0,1,1,1,0,1,1, the consecutive will look like 0,1,2,3,0,1,2 and then max consecutive will look like 0,3,0,2 or just 3,2 Jul 13 '20 at 12:27
• Iterating over dataframes is slow and usually not advised; it's better to look for an alternative. Jul 13 '20 at 12:45