diff options
Diffstat (limited to 'src')
| -rwxr-xr-x[-rw-r--r--] | src/cli.py | 29 | ||||
| -rwxr-xr-x[-rw-r--r--] | src/cost.py | 2 | ||||
| -rw-r--r-- | src/model.py | 17 | ||||
| -rwxr-xr-x[-rw-r--r--] | src/predict.py | 4 | ||||
| -rw-r--r-- | src/theta | 2 | ||||
| -rwxr-xr-x[-rw-r--r--] | src/train.py | 2 |
6 files changed, 32 insertions, 24 deletions
diff --git a/src/cli.py b/src/cli.py index ce759be..0c45e7a 100644..100755 --- a/src/cli.py +++ b/src/cli.py @@ -1,3 +1,5 @@ +#!/usr/bin/env python3.7 + import sys import argparse @@ -13,25 +15,25 @@ class CommandLineInterface: prog="ft_linear_regression_cli", description="CLI to interact with the ft_linear_regression project" ) - subparsers = parser.add_subparsers(help="sub-command help", dest="subparser_name") - + subparsers = parser.add_subparsers(dest="subparser_name") parser_train = subparsers.add_parser("train", help="train the model") parser_train.set_defaults(func=self._train) - parser_train.add_argument("-a --alpha", type=float, default=1.0, dest="alpha", help="learning rate") - parser_train.add_argument("-e --epoch", type=int, default=100, dest="epoch", help="number of iterations") - + parser_train.add_argument("-a --alpha", type=float, default=1.0, + dest="alpha", help="learning rate") + parser_train.add_argument("-e --epoch", type=int, default=100, + dest="epoch", help="number of iterations") parser_predict = subparsers.add_parser("predict", help="make a predict") parser_predict.set_defaults(func=self._predict) - parser_predict.add_argument("-x", type=int, help="mileage for which the prediction will be made") - + parser_predict.add_argument("-x", type=int, + help="mileage for which the prediction will be made") parser_cost = subparsers.add_parser("cost", help="print model cost") parser_cost.set_defaults(func=self._cost) - parser_plot = subparsers.add_parser("plot", help="plot data and model") parser_plot.set_defaults(func=self._plot) - parser_plot.add_argument("-d --data", help="only plot data", action="store_true", dest="plot_data") - parser_plot.add_argument("-m --model", help="only plot model", action="store_true", dest="plot_model") - + parser_plot.add_argument("-d --data", help="only plot data", + action="store_true", dest="plot_data") + parser_plot.add_argument("-m --model", help="only plot model", + action="store_true", dest="plot_model") self.args = parser.parse_args(sys.argv[1:]) def _train(self): @@ -40,7 +42,7 @@ class CommandLineInterface: def _predict(self): if self.args.x is not None: - print(self.model.hypothesis(self.args.x)) + print(self.model.make_prediction(self.args.x)) else: predict.predict_input(self.model) @@ -54,6 +56,9 @@ class CommandLineInterface: self.model.plot(self.args.plot_data, self.args.plot_model) def exec_args(self): + if self.args.subparser_name is None: + print("{} --help for more information".format(sys.argv[0])) + return self.args.func() diff --git a/src/cost.py b/src/cost.py index 0f00253..0f0743b 100644..100755 --- a/src/cost.py +++ b/src/cost.py @@ -1,3 +1,5 @@ +#!/usr/bin/env python3.7 + from model import Model diff --git a/src/model.py b/src/model.py index ee43050..0d74a69 100644 --- a/src/model.py +++ b/src/model.py @@ -1,6 +1,5 @@ import numpy as np import matplotlib.pyplot as plt -import sklearn.preprocessing class Model: @@ -25,6 +24,10 @@ class Model: def hypothesis(self, x): return x * self.theta1 + self.theta0 + def make_prediction(self, predict): + predict = (predict - self.xs.min()) / (self.xs.max() - self.xs.min()) + return self.hypothesis(predict) + def cost(self): return (1 / (2 * len(self.xs))) * sum([(self.hypothesis(x) - y) ** 2 for x, y in zip(self.xs, self.ys)]) @@ -42,8 +45,8 @@ class Model: def _plot_model(self): line_xs = [self.xs.min(), self.xs.max()] - line_ys = [self.hypothesis(x) for x in line_xs] - self.ax.plot(line_xs, line_ys) + line_ys = [self.make_prediction(x) for x in line_xs] + self.ax.plot(line_xs, line_ys, color='r') def _partial_theta1(self): return sum([(self.hypothesis(x) - y) * x @@ -54,7 +57,7 @@ class Model: for x, y in zip(self.xs, self.ys)]) / len(self.xs) def _normalize_data(self): - self.xs, self.ys = sklearn.preprocessing.normalize([self.xs, self.ys]) + self.xs = (self.xs - self.xs.min()) / (self.xs.max() - self.xs.min()) def _read_theta(self): try: @@ -72,9 +75,3 @@ class Model: return data[:, 0], data[:, 1] except IOError: print(self.datafilename, "do not exist") - - -if __name__ == "__main__": - m = Model() - m.train() - m.write_theta() diff --git a/src/predict.py b/src/predict.py index 329382a..2257109 100644..100755 --- a/src/predict.py +++ b/src/predict.py @@ -1,3 +1,5 @@ +#!/usr/bin/env python3.7 + from model import Model def predict_input(m): @@ -8,7 +10,7 @@ def predict_input(m): print("Bad input, you should enter a number") else: break - print("The predicted price for this mileage is", m.hypothesis(x)) + print("The predicted price for this mileage is", m.make_prediction(x)) if __name__ == "__main__": m = Model(thetafilename="./theta") @@ -1 +1 @@ --0.19808642684030997,0.23574835208002437
\ No newline at end of file +-4615.249629014704,7992.7771682847515
\ No newline at end of file diff --git a/src/train.py b/src/train.py index 0a68916..146d0ee 100644..100755 --- a/src/train.py +++ b/src/train.py @@ -1,3 +1,5 @@ +#!/usr/bin/env python3.7 + from model import Model |