Added Gradient Descent [Python] #348

Prakash2403 · 2017-06-22T09:58:39Z

Fixes #345

By submitting this pull request I confirm I've read and complied with the below declarations.

I have read the Contribution guidelines and I am confident that my PR reflects them.
I have followed the coding guidelines for this project.
My code follows the skeleton code structure.
This pull request has a descriptive title. For example, Added {Algorithm/DS name} [{Language}], not Update README.md or Added new code.
This pull request will be closed if I fail to update it even once in a continuous time span of 7 days.

yashLadha

Please see comments

yashLadha · 2017-06-22T10:19:54Z

gradient_descent/gradient_descent.py

+    :param data_input_tuple: Input tuple of a particular example
+    :return: Value of hypothesis function at that point.
+    Note that parameter input value is fixed as 1.
+    Also known as 'biased input' inn ML terminology and the parameter associated with it


change inn to in

Prakash2403 · 2017-06-22T11:47:37Z

@yashLadha Please review again.

aashutoshrathi · 2017-06-22T12:50:38Z

gradient_descent/gradient_descent.py

+    while True:
+        j = j+1
+        temp_parameter_vector = [0, 0, 0, 0]
+        for i in range(0, len(parameter_vector)):


for i in range(0, len(parameter_vector)): can be changed to for i in range(len(parameter_vector)):

aashutoshrathi · 2017-06-22T12:51:26Z

gradient_descent/gradient_descent.py

+
+def test_gradient_descent():
+    for i in range(len(test_data)):
+        print("Actual output value:",  output(i, 'test'))


print("Actual output value:", output(i, 'test')) should be changed to print("Actual output value:", output(i, 'test')) i.e. remove one space between , and output..

aashutoshrathi · 2017-06-22T13:01:45Z

gradient_descent/gradient_descent.py

+    relative_error_limit = 0
+    j = 0
+    while True:
+        j = j+1


can be changed to j += 1

Prakash2403 · 2017-06-22T13:09:34Z

@aashutoshrathi Please review again.

aashutoshrathi · 2017-06-22T13:12:40Z

gradient_descent/gradient_descent.py

+    hyp_val = 0
+    for i in range(len(parameter_vector) - 1):
+        hyp_val = hyp_val + data_input_tuple[i]*parameter_vector[i+1]
+    hyp_val = hyp_val + 1*parameter_vector[0]


@Prakash2403 Sir , Why is this 1*parameter_vector[0] not parameter_vector[0]?

Line 35 and 36 can be changed to += format.

As I have mentioned in comments, there is always a biased input in Artificial Neural Networks or any ML hypothesis, whose value is fixed as 1. I wanted to explicitly mention this fact in the code too, that's why I have written 1*parameter_vector[0] instead of parameter[0].

But now, I guess it's better to keep it parameter[0]. Thanks for pointing it out.

Prakash2403 · 2017-06-22T13:35:14Z

@aashutoshrathi Please review again.

aashutoshrathi · 2017-06-22T14:31:42Z

gradient_descent/gradient_descent.py

+    :return: Value of hypothesis function at that point.
+    Note that there is an 'biased input' whose value is fixed as 1.
+    It is not explicitly mentioned in input data.. But, ML hypothesis functions use it.
+    So, we have to take care of it separately. Line 36 takes care of it. 


Remove Trailing Whitespaces

Where is the issue exactly?

At the end of Line 31

Got it, thanks.

Prakash2403 · 2017-06-22T16:29:20Z

@aashutoshrathi Please review again.

yashLadha · 2017-06-25T11:49:34Z

Just update your branch

- Added open source datasets

yashLadha

Please See comments

yashLadha · 2017-06-27T12:09:57Z

gradient_descent/gradient_descent.py

+    plt.ylabel('Output Values')
+    plt.xlim([-1, 7])
+    plt.legend()
+    plt.show()


Won't the program end. If u haven't close the plot explicitly. May be this will hold the travis-ci build to complete. Make a timer or something so that it get closed.

yashLadha

Please See Comments

yashLadha · 2017-06-27T18:29:26Z

gradient_descent/gradient_descent.py

    parameter_vector = numpy.asmatrix([2, 4, 1, 5, 4, 1, 2, 2, 3, 1, 1, 2]).transpose()
    learning_rate = 0.00015
-    absolute_error_limit = 0.000015
+    absolute_error_limit = 0.000015/4


Why this is so? Why you are not using the simplified version?.

yashLadha · 2017-06-27T18:31:30Z

gradient_descent/gradient_descent.py

+    """
+    Downloads test and train data from GitHub repository
+    """
+    import requests


Put the import , where all the declarations are defined

yashLadha · 2017-06-28T05:34:59Z

There are conflicts in the README.md please resolve them.

Prakash2403 · 2017-06-28T05:40:09Z

@yashLadha Please review again.

yashLadha · 2017-06-28T06:51:55Z

@Prakash2403 have u passed a parameter for plotting the graph. ??

yashLadha

Please see comments

yashLadha · 2017-06-28T06:52:49Z

gradient_descent/gradient_descent.py

+    hypothesis_output = _hypothesis_value(input_data,
+                                          parameter_vector=parameter_vector)
+    num_examples = len(output_data)
+    plt.plot(range(num_examples), actual_output, 'r', label='Actual Output')


Check for argument passed for plotting the graph

yashLadha · 2017-06-28T06:54:26Z

Add matplotlib to requirements. @Prakash2403

- Added matplotlib in python3 requirements.txt

Prakash2403 · 2017-06-28T17:38:04Z

@yashLadha Please review again.

yashLadha

Nide Work 😄 👍

yashLadha · 2017-06-28T18:54:13Z

gradient_descent/gradient_descent.py

+    actual_output = output_data
+    hypothesis_output = _hypothesis_value(input_data,
+                                          parameter_vector=parameter_vector)
+    if len(sys.argv) == 2:


aviaryan · 2017-06-30T03:50:03Z

@Prakash2403 Can you please rebase this PR? matplotlib has been already added.

Prakash2403 · 2017-06-30T03:56:44Z

@aviaryan Sir, there is an on-going discussion regarding this issue on gitter. So, it is best to not merge the PR right now.

Monal5031 · 2017-08-18T15:58:20Z

@Prakash2403 sir please update this PR.

singhpratyush · 2017-09-10T18:52:23Z

Please reopen if you wish to continue.

= added 2 commits June 22, 2017 15:21

- Added Gradient Descent

a64e36d

- Fixed typo

84c1a6d

Prakash2403 requested review from aviaryan, singhpratyush and yashLadha June 22, 2017 09:58

Prakash Rai added 2 commits June 22, 2017 15:29

Merge branch 'master' into master

e7246c4

- Changed gradient_descent.py according to PEP8 guidelines.

9ddb8b9

yashLadha reviewed Jun 22, 2017

View reviewed changes

Fixed typos in comments

f3cb691

aashutoshrathi reviewed Jun 22, 2017

View reviewed changes

gradient_descent/gradient_descent.py Outdated

relative_error_limit = 0

j = 0

while True:

j = j+1

Copy link

Member

aashutoshrathi Jun 22, 2017

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can be changed to j += 1

Update gradient_descent.py

3f7c2d1

aashutoshrathi reviewed Jun 22, 2017

View reviewed changes

Update gradient_descent.py

29ce52d

aashutoshrathi approved these changes Jun 22, 2017 •

edited

Loading

View reviewed changes

aashutoshrathi reviewed Jun 22, 2017

View reviewed changes

Prakash Rai added 2 commits June 22, 2017 21:09

Merge branch 'master' into master

d88fe72

Update gradient_descent.py

3556dc3

aashutoshrathi approved these changes Jun 22, 2017

View reviewed changes

Merge branch 'master' into master

a2d3d46

yashLadha approved these changes Jun 25, 2017

View reviewed changes

Prakash Rai added 3 commits June 25, 2017 17:39

Merge branch 'master' into master

c4f79bf

Merge branch 'master' into master

f0a6367

- Modified gradient_descent.py

77c58df

- Added open source datasets

yashLadha requested changes Jun 27, 2017

View reviewed changes

- Updated gradient_descent.py

6da9dab

yashLadha requested changes Jun 27, 2017

View reviewed changes

yashLadha reviewed Jun 27, 2017

View reviewed changes

Update gradient_descent.py

c3d6150

Merge branch 'master' into master

c74af4b

yashLadha requested changes Jun 28, 2017

View reviewed changes

- Graph is shown when sys argument is passed.

3802026

- Added matplotlib in python3 requirements.txt

Prakash Rai added 3 commits June 28, 2017 23:17

Update gradient_descent.py

7a0b0c9

Update pip2-requirements.txt

5eea0f9

Merge branch 'master' into master

601d0fb

yashLadha reviewed Jun 28, 2017

View reviewed changes

yashLadha approved these changes Jun 28, 2017

View reviewed changes

Merge branch 'master' into master

d592d32

Monal5031 assigned Prakash2403 Jun 30, 2017

Monal5031 added Algorithm lang-python labels Jun 30, 2017

singhpratyush assigned aviaryan and unassigned Prakash2403 Jul 1, 2017

singhpratyush added the No activity label Sep 8, 2017

singhpratyush closed this Sep 10, 2017

Added Gradient Descent [Python] #348

Added Gradient Descent [Python] #348

Conversation

Prakash2403 commented Jun 22, 2017

By submitting this pull request I confirm I've read and complied with the below declarations.

yashLadha left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Prakash2403 commented Jun 22, 2017

Choose a reason for hiding this comment

aashutoshrathi Jun 22, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Prakash2403 commented Jun 22, 2017

aashutoshrathi Jun 22, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Prakash2403 Jun 22, 2017 • edited Loading

Choose a reason for hiding this comment

Prakash2403 commented Jun 22, 2017

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Prakash2403 commented Jun 22, 2017

yashLadha commented Jun 25, 2017

yashLadha left a comment

Choose a reason for hiding this comment

yashLadha Jun 27, 2017 • edited Loading

Choose a reason for hiding this comment

yashLadha left a comment

Choose a reason for hiding this comment

yashLadha Jun 27, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

yashLadha commented Jun 28, 2017

Prakash2403 commented Jun 28, 2017

yashLadha commented Jun 28, 2017

yashLadha left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

yashLadha commented Jun 28, 2017

Prakash2403 commented Jun 28, 2017

yashLadha left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

aviaryan commented Jun 30, 2017

Prakash2403 commented Jun 30, 2017

Monal5031 commented Aug 18, 2017

singhpratyush commented Sep 10, 2017

aashutoshrathi Jun 22, 2017 •

edited

Loading

aashutoshrathi Jun 22, 2017 •

edited

Loading

Prakash2403 Jun 22, 2017 •

edited

Loading

yashLadha Jun 27, 2017 •

edited

Loading

yashLadha Jun 27, 2017 •

edited

Loading