KaggleStats Module

Some statistics (printed on the screen) about the datas.


Sortie du script

$ python KaggleStats.py
OK, I read this file, I have now a huge array (of size : (891, 11)) !
This array have 891 passengers, and the following attributes :
 the 0th attribute is survived.
 the values of this attribute are known for 100.00% of the passengers
 the 1th attribute is pclass.
 the values of this attribute are known for 100.00% of the passengers
 the 2th attribute is name.
 the values of this attribute are known for 100.00% of the passengers
 the 3th attribute is sex.
 the values of this attribute are known for 100.00% of the passengers
 the 4th attribute is age.
 the values of this attribute are known for 80.13% of the passengers
 the 5th attribute is sibsp.
 the values of this attribute are known for 100.00% of the passengers
 the 6th attribute is parch.
 the values of this attribute are known for 100.00% of the passengers
 the 7th attribute is ticket.
 the values of this attribute are known for 100.00% of the passengers
 the 8th attribute is fare.
 the values of this attribute are known for 100.00% of the passengers
 the 9th attribute is cabin.
 the values of this attribute are known for 22.90% of the passengers
 the 10th attribute is embarked.
 the values of this attribute are known for 99.78% of the passengers
This array contains 891 passengers...
and 342 passengers survived the Titanic accident in 1912 :(
(so 549 passengers died)
which gives a proportion of 0.3838 survivors.

Considering the women and men ...
The training data set contains 314 women,
The training data set contains 577 men.
Proportion of women who survived is 0.7420,
Proportion of men who survived is 0.1889.

Considering the three different embarcations points:
 * S = Southampton (EN) -> 1;
 * C = Cherbourg (FR) -> 0;
 * Q = Queenstown (EN) -> 2.
 
644 passengers took the Titanic at Southampton;
168 passengers took the Titanic at Cherbourg;
77 passengers took the Titanic at Queenstown.

Proportion of survivors, for the Southampton part, is 0.3370;
Proportion of survivors, for the Cherbourg part, is 0.5536;
Proportion of survivors, for the Queenstown part, is 0.3896.

Considering the age of the passengers...
The age is known for 80.13% of the passengers (so 714 passengers).
For them, the min age is 0.42.
For them, the max age is 80.00.
For them, the median age is 29.70.
The median age of the victims is 30.626.
The median age of the survivors is 28.344.
... so : no obvious link between age and survivance rate ?
Yeah, but maybe median age is not the best way to show it.

Considering the fare of the passengers...
The fare is known for 100.00% of the passengers (so 891 passengers).
For them, the min fare is 0.00.
For them, the max fare is 512.33.
For them, the median fare is 32.20.
The median fare of the victims is 22.118.
The median fare of the survivors is 48.395.
Opening the file 'test.csv'...
This array have 418 passengers, and the following attributes :
 the 0th attribute is pclass.
   the values of this attribute are known for 100.00% of the passengers
 the 1th attribute is name.
   the values of this attribute are known for 100.00% of the passengers
 the 2th attribute is sex.
   the values of this attribute are known for 100.00% of the passengers
 the 3th attribute is age.
   the values of this attribute are known for 79.43% of the passengers
 the 4th attribute is sibsp.
   the values of this attribute are known for 100.00% of the passengers
 the 5th attribute is parch.
   the values of this attribute are known for 100.00% of the passengers
 the 6th attribute is ticket.
   the values of this attribute are known for 100.00% of the passengers
 the 7th attribute is fare.
   the values of this attribute are known for 99.76% of the passengers
 the 8th attribute is cabin.
   the values of this attribute are known for 21.77% of the passengers
 the 9th attribute is embarked.
   the values of this attribute are known for 100.00% of the passengers

Table des matières

Cette page en .txt et en .pdf

Sujet précédent

KaggleModel Module

Sujet suivant

RandomForest Module