def index():
if request.method == "POST":
file_content = flask.request.files.get('file_content', '')
if not file_content:
prediction = "select an image!!"
else:
filename = secure_filename(file_content.filename)
save_path = os.path.join("faceorplace", "static", "undecided",
filename)
file_content.save(save_path)
prediction = model.make_prediction(file_content)
move_path = os.path.join("faceorplace", "static", prediction + "s",
filename)
os.rename(save_path, move_path)
time.sleep(0.1)
if request.method == "GET":
prediction = "..."
imgs_faces = get_image_files(os.path.join("faceorplace", "static",
"faces"))
imgs_faces = ["static/faces/" + f for f in imgs_faces]
imgs_places = get_image_files(
os.path.join("faceorplace", "static", "places"))
imgs_places = ["static/places/" + f for f in imgs_places]
return render_template('index.jinja2',
imgs_faces=imgs_faces,
imgs_places=imgs_places,
prediction=prediction)
def model_prediction():
# We retrieve the data payload of the POST request
data = request.get_json(force=True)
# We then preprocess our data, and use our pretrained model to make a
# prediction.
output = make_prediction(data, static_model)
# We finally package this prediction as a JSON object to deliver a valid
# response with our API.
return jsonify(output)
def main(train_file, test_file, model_path, output_path, plot_flag):
if train_file:
click.echo('Loading training data')
X_train, y_train = load_data(train_file, model_path)
click.echo('Generating corr heatmap')
if plot_flag:
try:
plot_corr(X_train, y_train, output_path, plot_it=False)
except:
click.echo('Failed to plot correlation heatmap')
click.echo('Model traninig')
train_xgbr(X_train,
y_train,
INITIAL_PARAM,
PARAM_1,
PARAM_2,
model_path=model_path,
test_size=0.2)
click.echo('Loading test data')
X_test, y_test = load_data(test_file, model_path)
click.echo('Making prediction')
pred = make_prediction(X_test, model_path, output_path)
click.echo('Evaluating and plot')
_, scores = evaluate(pred,
y_test,
X_test,
output_path,
window_size=WIN_SIZE,
top=TOP,
ranks=RANK)
def model_prediction():
file_name = flask.request.values.get('file_name')
return model.make_prediction(file_name)
import csv
import model
# make a prediction given the data
predictions = []
with open('./train.csv') as data_file:
data = data_file.read()
for row in data:
predictions.append(model.make_prediction(row))
# write the data to a file to submit
fieldnames = ['PassengerId', 'Survived']
with open('predictions.csv', 'w+') as csv_file:
writer = csv.DictWriter(csv_file, fieldnames=fieldnames)
writer.writeheader()
for prediction in predictions:
writer.writerow(prediction)
