CODEWITHSUNDEEP
pythonpandas
The Nightman
The Nightman

Reputation: 5759

Pandas mean of list within dataframe

I have a pandas DataFrame that has column which contains lists. I am trying to get the means of the lists in this column.

Here is an example of what my DataFrame looks like:

    Loc         Background
0   115227854   [0.000120481927711]
1   115227854   [0.000129117642312, 0.000131429072111, 0.00016...
2   115227855   [0.000123193166886]
3   115227855   [0.000142845482001, 0.000184789750329, 0.00018...
4   115227856   [0.000173490631506]

I would like to do something like this to set a new Mean column equal to the mean of the data in each of the lists found in the Background column:

sig_vars['Mean'] = sig_vars['Background'].mean()

And here is the DataFrame if needed:

df = {'Background': {0: [0.00012048192771084337],
  1: [0.00012911764231185137,
   0.0001314290721107509,
   0.000163015792154865,
   0.00018832391713747646,
   0.00019627513412134165,
   0.00020383723596708027,
   0.0002114408734430263,
   0.00022564565426983117,
   0.000247843759294141],
  2: [0.00012319316688567673],
  3: [0.00014284548200146926,
   0.00018478975032851512,
   0.00018864365214110544,
   0.00019392685725367248,
   0.00022931689046296532,
   0.00023965141612200435,
   0.00036566589684372596,
   0.00043096760847454704,
   0.0004584752423369138],
  4: [0.00017349063150589867]},
 'Loc': {0: 115227854, 1: 115227854, 2: 115227855, 3: 115227855, 4: 115227856}}

Upvotes: 2

Views: 15258

Answers (5)

jbb
jbb

Reputation: 76

Use pandas.Series.apply:

    df['Mean'] = df['Background'].apply(np.mean)

Upvotes: 3

shubham_kamath
shubham_kamath

Reputation: 389

Here is what I can think of.

  1. Iterate through the specific column and and store it's mean in a DataFrame.

    df = pandas.DataFrame(sig_vars.iloc[i]['background'].mean() for i in range(len(sig_vars)),columns=['mean'])

  2. Join the column with the main dataframe.

    sig_vars = sig_vars.join(df)

Upvotes: 1

b2002
b2002

Reputation: 914

list comprehension converting each list to array

df['Mean'] = [np.array(x).mean() for x in df.Background.values]

Upvotes: 1

Rahul Agarwal
Rahul Agarwal

Reputation: 4100

Use can also use np.mean to achieve the same:

import numpy as np
np.mean(df['Background'].tolist(), axis=1)

Upvotes: 5

BENY
BENY

Reputation: 323276

Using tolist recreate the dataframe

pd.DataFrame(sig_vars['Background'].values.tolist()).mean(1)
Out[498]: 
0    0.000120
1    0.000189
2    0.000123
3    0.000270
4    0.000173
dtype: float64

#sig_vars['Mean'] = pd.DataFrame(sig_vars['Background'].values.tolist()).mean(1)

Upvotes: 4

Related Questions