partial fitting, how to ensure one hot captures all features consistently
$begingroup$
Doing some data science on ~4 million samples, with lots of columns being categorical.
One column has ~1000 categories and my boss insists on including it in the analysis.
My output is also predicting classes (I'll use gnb.predict_proba())
So, I'm taking a random subset of my data for partial fitting, and repeating.
# train = ~3 million rows of data as a dataframe
gnb = naive_bayes.GaussianNB()
for i in range(10):
dds = train.sample(n=10**4)
(dfX,dfY) = makeXY(dds) #gets one-hot- encoded X and Y dataframes
gnb.partial_fit(dfX,[getClass(x) for x in dfY.values],classes=np.unique([getClass(x) for x in dfY.values]))
How can I ensure I get all the possible classes AND that they are in the same order every time?
python pandas
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
$endgroup$
bumped to the homepage by Community♦ 18 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
add a comment |
$begingroup$
Doing some data science on ~4 million samples, with lots of columns being categorical.
One column has ~1000 categories and my boss insists on including it in the analysis.
My output is also predicting classes (I'll use gnb.predict_proba())
So, I'm taking a random subset of my data for partial fitting, and repeating.
# train = ~3 million rows of data as a dataframe
gnb = naive_bayes.GaussianNB()
for i in range(10):
dds = train.sample(n=10**4)
(dfX,dfY) = makeXY(dds) #gets one-hot- encoded X and Y dataframes
gnb.partial_fit(dfX,[getClass(x) for x in dfY.values],classes=np.unique([getClass(x) for x in dfY.values]))
How can I ensure I get all the possible classes AND that they are in the same order every time?
python pandas
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
$endgroup$
bumped to the homepage by Community♦ 18 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
$begingroup$
Look here for stratified sampling: stackoverflow.com/questions/41035187/…
$endgroup$
– CrazyElf
Jan 31 '18 at 14:33
add a comment |
$begingroup$
Doing some data science on ~4 million samples, with lots of columns being categorical.
One column has ~1000 categories and my boss insists on including it in the analysis.
My output is also predicting classes (I'll use gnb.predict_proba())
So, I'm taking a random subset of my data for partial fitting, and repeating.
# train = ~3 million rows of data as a dataframe
gnb = naive_bayes.GaussianNB()
for i in range(10):
dds = train.sample(n=10**4)
(dfX,dfY) = makeXY(dds) #gets one-hot- encoded X and Y dataframes
gnb.partial_fit(dfX,[getClass(x) for x in dfY.values],classes=np.unique([getClass(x) for x in dfY.values]))
How can I ensure I get all the possible classes AND that they are in the same order every time?
python pandas
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
$endgroup$
Doing some data science on ~4 million samples, with lots of columns being categorical.
One column has ~1000 categories and my boss insists on including it in the analysis.
My output is also predicting classes (I'll use gnb.predict_proba())
So, I'm taking a random subset of my data for partial fitting, and repeating.
# train = ~3 million rows of data as a dataframe
gnb = naive_bayes.GaussianNB()
for i in range(10):
dds = train.sample(n=10**4)
(dfX,dfY) = makeXY(dds) #gets one-hot- encoded X and Y dataframes
gnb.partial_fit(dfX,[getClass(x) for x in dfY.values],classes=np.unique([getClass(x) for x in dfY.values]))
How can I ensure I get all the possible classes AND that they are in the same order every time?
python pandas
python pandas
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
edited Jan 31 '18 at 14:28
Stephen Rauch♦
1,53551330
1,53551330
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
asked Jan 31 '18 at 14:02
Hari PrasadHari Prasad
126124
126124
bumped to the homepage by Community♦ 18 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
bumped to the homepage by Community♦ 18 mins ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
$begingroup$
Look here for stratified sampling: stackoverflow.com/questions/41035187/…
$endgroup$
– CrazyElf
Jan 31 '18 at 14:33
add a comment |
$begingroup$
Look here for stratified sampling: stackoverflow.com/questions/41035187/…
$endgroup$
– CrazyElf
Jan 31 '18 at 14:33
$begingroup$
Look here for stratified sampling: stackoverflow.com/questions/41035187/…
$endgroup$
– CrazyElf
Jan 31 '18 at 14:33
$begingroup$
Look here for stratified sampling: stackoverflow.com/questions/41035187/…
$endgroup$
– CrazyElf
Jan 31 '18 at 14:33
add a comment |
1 Answer
1
active
oldest
votes
$begingroup$
If you're using sklearn, this is a great use of CountVectorizer as a workaround since you can specify a vocabulary.
To start, get a list of all the 1000 categories and set that as the vocabulary in the transformer. Then convert the column to a string data type and apply this transformer to each batch:
from sklearn.feature_extraction.text import CountVectorizer
cv = CountVectorizer(vocabulary = pre_made_category_list)
endoded_variable_matrix = cv.fit_transform(dfX[categorical_column])
Even though it's technically a count vectorizor, since there is only one word in the string, the counts for each row will be 1 for the category it is and 0 for everything else, so its effectively one hot encoding the variable. The order of the columns will be the order of the vocabulary, so the matrix will be consistent between folds.
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
$endgroup$
add a comment |
Your Answer
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "557"
};
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function() {
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled) {
StackExchange.using("snippets", function() {
createEditor();
});
}
else {
createEditor();
}
});
function createEditor() {
StackExchange.prepareEditor({
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: false,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: null,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f27297%2fpartial-fitting-how-to-ensure-one-hot-captures-all-features-consistently%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
If you're using sklearn, this is a great use of CountVectorizer as a workaround since you can specify a vocabulary.
To start, get a list of all the 1000 categories and set that as the vocabulary in the transformer. Then convert the column to a string data type and apply this transformer to each batch:
from sklearn.feature_extraction.text import CountVectorizer
cv = CountVectorizer(vocabulary = pre_made_category_list)
endoded_variable_matrix = cv.fit_transform(dfX[categorical_column])
Even though it's technically a count vectorizor, since there is only one word in the string, the counts for each row will be 1 for the category it is and 0 for everything else, so its effectively one hot encoding the variable. The order of the columns will be the order of the vocabulary, so the matrix will be consistent between folds.
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
$endgroup$
add a comment |
$begingroup$
If you're using sklearn, this is a great use of CountVectorizer as a workaround since you can specify a vocabulary.
To start, get a list of all the 1000 categories and set that as the vocabulary in the transformer. Then convert the column to a string data type and apply this transformer to each batch:
from sklearn.feature_extraction.text import CountVectorizer
cv = CountVectorizer(vocabulary = pre_made_category_list)
endoded_variable_matrix = cv.fit_transform(dfX[categorical_column])
Even though it's technically a count vectorizor, since there is only one word in the string, the counts for each row will be 1 for the category it is and 0 for everything else, so its effectively one hot encoding the variable. The order of the columns will be the order of the vocabulary, so the matrix will be consistent between folds.
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
$endgroup$
add a comment |
$begingroup$
If you're using sklearn, this is a great use of CountVectorizer as a workaround since you can specify a vocabulary.
To start, get a list of all the 1000 categories and set that as the vocabulary in the transformer. Then convert the column to a string data type and apply this transformer to each batch:
from sklearn.feature_extraction.text import CountVectorizer
cv = CountVectorizer(vocabulary = pre_made_category_list)
endoded_variable_matrix = cv.fit_transform(dfX[categorical_column])
Even though it's technically a count vectorizor, since there is only one word in the string, the counts for each row will be 1 for the category it is and 0 for everything else, so its effectively one hot encoding the variable. The order of the columns will be the order of the vocabulary, so the matrix will be consistent between folds.
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
$endgroup$
If you're using sklearn, this is a great use of CountVectorizer as a workaround since you can specify a vocabulary.
To start, get a list of all the 1000 categories and set that as the vocabulary in the transformer. Then convert the column to a string data type and apply this transformer to each batch:
from sklearn.feature_extraction.text import CountVectorizer
cv = CountVectorizer(vocabulary = pre_made_category_list)
endoded_variable_matrix = cv.fit_transform(dfX[categorical_column])
Even though it's technically a count vectorizor, since there is only one word in the string, the counts for each row will be 1 for the category it is and 0 for everything else, so its effectively one hot encoding the variable. The order of the columns will be the order of the vocabulary, so the matrix will be consistent between folds.
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
answered Jan 31 '18 at 22:32
dbagherndbaghern
29115
29115
add a comment |
add a comment |
Thanks for contributing an answer to Data Science Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f27297%2fpartial-fitting-how-to-ensure-one-hot-captures-all-features-consistently%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
$begingroup$
Look here for stratified sampling: stackoverflow.com/questions/41035187/…
$endgroup$
– CrazyElf
Jan 31 '18 at 14:33