cs340-21 / mathmajorsunite Goto Github PK

The HTML of the process_text.html file needs to be cleaned up. If we could make this an appealing layout, that would be very preferable because this page is the bulk of the website.

ESKETIT

Bug currently in image.html code

Add some information about the project on the page "About" on the website.

Find theme that we can use in Bootstrap to improve the appearance of the site.

Check in Image and Text uploading if the files are too large (~10MB or so), and reject the upload if so.

On the backend, we need some efficient way to merge CSV files and save them locally. This function should be able to do the following:

Arguments:

1. list of pandas dataframes

Returns:
-1 if files were not able to be merged
0 if files were able to be merged

This function should be able to check if the files are able to be merged. Just try to indicate if there is an error or not in the return, I can show an error message on the front end if needed.

It should also be able to save the merged dataframe to the location from which the other dataframes were loaded. If you want to, you could return the merged dataframe and I could save it from the frontend.

Currently, the HTML files don't have the best redirection between them (i.e. it's not easy to navigate between pages). If we could improve the linkage between the pages (possibly putting in a sidebar or something), that would contribute to the overall look of the site.

I need a function that will plot a histogram for a column in a dataframe. This will contribute to our reporting functionality in the website. The function should make the histogram and save it to a folder, and it should return the absolute path to that image.

The function should take the following arguments:
dataframe: pandas dataframe
column: name of column on which to make the histogram

The function should return:
-1 if error (will be needed for processing)
1 if the data is categorical
image file name (string) if successful: should be the name of the file (with absolute path) where the image of the histogram has been saved

Establish system for handling files in the website so that the backend can easily access them.

Functionality to normalize the columns in the pandas dataframe.

Add more information to proposal, including:
Testing our product
Specific team roles
More detailed schedule

Set up table to rename columns - extract data for function call.

Get some buttons/functions set up to run the processing

Lay out a plan for calling processing functions in the main scripts.

For every function that edits a dataframe, I need the functions on the backend to save the CSV's back to exactly where they came from (i.e. the same location that they had before they were loaded). This will involve multiple functions, just whichever ones edit the dataframes.

All files loaded in from the website are stored in the MathMajorsUnite/cs340_project/media/media/text directory. I can pass the path from which I loaded the CSV file, whatever is easier. Please let me know.

Need to have base sample data for preliminary tests.

I am just creating the issue now, so that I can edit it later.

I noticed that in the read.py backend file, a lot of the functionality is implemented throughout command-line interaction (using input() function to extract from stdin). For the purposes of the frontend, we need to remove the input() control flow of these functions and convert this to an arguments-only framework. For example, in read_files in the Read class, we need the merging functionality to be able to be activated by a function call, avoiding any processing of stdin.

Make the linear regression page more interpretable, like put in some descriptions and/or links to resources.

On visualize.html, the drop down menus could definitely use some visualization improvement. Good links:

https://www.w3schools.com/howto/howto_css_dropdown.asp
https://www.w3schools.com/howto/howto_custom_select.asp

Implement PCA, TSNE 2d visualization for the data. Should have the following features:

1. Choose which column to use as label
2. Show labels in legend
3. Grab pandas from front-end, send to backend and perform dim reduction (sklearn, .to_numpy())

How to display plots, visual data in the front end.

First rough draft of the Project Proposal