Tested 3–4 Issues

I had go through almost all the issues of the tracker and trainer app. I had tested issues 4 issues marked with testing label and prepared their solution. I will be discussing my approaches for these issues with my mentor and fixed them in next week.

Fixed mhbs Logo Issue Issue #4

In the tracker app some of the screens have the dhis-2 logo. It is now fixed to mhbs logo.

Fixed Issue #4 

Related MR : Mhbs logo fixed

Updated Trainer App Readme.md

Since we had added media page, offline support, app usage tracker in trainer app we want a small documentation details to be added to the readme for further reference. I had done it -

Committed here ( 718f8756 ).

Fixed Bug in GitLab CI

There is a bug in GitLab CI code of trainer app. Gradle is building the apk correctly but it is not assembling the artifact correctly. It is my mistake what I had made earlier. I had corrected it.

Committed f01c6cf9.

Work Done this Week

• Fixed logo Issue. 
• Updated Trainer App Readme.md.
• Fixed Bug in GitLab CI.
• Tested 3–4 Issues & prepared their solution. After discussing with mentor need to implement them.

Planning of next Week

• Need to modify HBB infrastructure survey
• Need to change frequency of tracker forms and HBB survey
• Fixing more issues which I had tested this week and prepared solution to fix them.

The video object detection pipeline is designed using openCV. The videos are loaded and a frame rate of 25FPS is assumed. Using the frames the tf.session is initiated and TF Object detection API model is run. The resulting bounding boxes are drawn using openCV.

Resulting images with bounding boxes are written to video with 1FPS using openCV videowriter. After that the generated video can be served using data or blob url depending upon the video so generated.

The segmentation model is not working properly for most of the cases. The obvious solution for it would be to retrain the entire model. The segmentation model has to be implemented in the same way as the Object detection using openCV.

At present the entire video generation pipeline takes less than 40 seconds. This excludes the uploading time and downloading time. The server on which it was tested has a RTX 3060 GPU peak GPU memory utilisation reached upto 11GB

That's all for today!!!

Hope you had a great week

Annyeong

 Week 7 Coding period GSOC 2021 LibreHealth-

Hey, Reader's we are in the 17th week of the coding period.

The 7th week started on 16th July. This week I was working on web application nearby hospital feature.

What progress have you made this week-

In the 7th week of the coding period, I have done the following thing with the cost of care web application.

1-Added Nearby Hospital to the Home screen of web app using the overpass api.

2- UI change of Home Screen.

3- added search by address and name filter to hospital.

4- sent 1 MR.

link - https://gitlab.com/librehealth/toolkit/cost-of-care/lh-toolkit-cost-of-care-explorer/-/merge_requests/12

My plan for next week- I have planned to do the following task for week 8

1-UI build for compare hospital with star rating.

2-Add compare Hospital Screen flutter app.

I enjoyed working on this feature and I am excited about next week's task.

See you next week.

Week 7 Of coding period is completed, I worked on adding on call doctors viewing feature, I completed notifcation showing functionality on risk assessments and monitoring alerts

Added On-Call Doctors Viewing Functionality in App with Unit tests

• Fetching events from on call doctors program
• displaying on call doctors
• displaying images with cache
• saving the list to local storage

Unit tests

• yield OnCallDoctorLoaded on successful fetching of data
• yields OnCallDoctorLoaded on unsuccessful fetching of data by getting from hive
• returns response with 200 code on successful fetching of on call doctors
• throws fetch data exception on no internet connection

GIF Showing on Call Doctors slider

Completed Notification screen (Risk Assessment, Monitoring Alerts Tab) with Unit tests

• Created notification screen bloc
• Created repository with methods to get risk assessments and monitoring alerts
• on tap of the list item the child assessments are opened

Unit tests

• Flutter Notifcation BloC test
• Test Cases
• yield NotificationLoaded on Fetch Nofitication Of Babies
• Flutter Notifcation Repository test
• Test Cases
• Risk Assessments Test cases
• status changed normal to danger
• status changed problem to danger
• status changed normal to problem
• status changed danger to problem
• Monitoring Alerts Test cases
• stage-4 assessments next

Unit tests Phase 4

AssessmentsRepository Test Cases

• Throws exception if assesments are done before 180 mins from birth
• Throws exception if assessments are incomplete
• Marks phase-4 as complete if everything is correctly filled
• Adds stage-4 assessments if list is empty

AssessmentsBloc Test Cases

• yields AssessmentsAdded on successful adding of phase 4
• yields AssessmentsError when assessments are incomplete while adding of phase 4 assessments

Phase 3 Assessments Unit testing

Test Cases

• Throws exception if Stage3 Normal assessments are incomplete
• Throws exception if Stage3 Problem assessments are incomplete
• Throws exception if Stage3 Danger assessments are incomplete
• Marks Stage3 Normal as complete if everything is correctly filled
• Marks Stage3 Problem as complete if everything is correctly filled
• Marks Stage3 Danger as complete if everything is correctly filled
• Adds Stage3 Normal assessments if list size is 2 and classification is normal
• Adds Stage3 Problem assessments if list size is 2 and classification is problem
• Adds Stage3 Danger assessments if list size is 2 and classification is danger

Bloc Test

• yields AssessmentsAdded on successful adding of phase 3
• yields AssessmentsError when assessments are incomplete while adding of phase 3 assessments

What progress have you made this week?

• Added On-Call Doctors Viewing Functionality in App
• Completed Notification screen (Risk Assessment, Monitoring Alerts Tab)
• Phase - 3,4 Unit testing

What do I plan to do next week?

1. Fetching & displaying doctor’s schedule in App with unit tests
2. Making messaging/Notifications viewing functionality in App with unit tests
3. Creating share App report bug features
4. Creating Stage 5 (Discharge) Stage with unit tests

PROGRESS MADE THIS WEEK:

Hi, According to the proposal timeline, week 6 was to be based on:

• finishing pending work.
• fixing bugs.
• starting week7 tasks.

Fortunately, there was no pending work, hence the focus was mainly on the remaining points.

The tasks of week7 are to:

• implement the visit-nurse and visit-provider workflow in the react application.

As most of the components required for implementing the visit-nurse workflow are already present in the web-component library, I started working on it.

DETAILED DESCRIPTION:

The visit-nurse workflow consists of many screens and multiple components.

• First screen contains a find-patient component to select a patient.
• After the patient is selected a redux action is triggered which updates redux state and stores the selected patient Id.
• Another redux action updates the active screen and changes it to that of patient encounters, it also updates redux state and stores the current encounter Id.
• The patient encounter screens contains all the patient appointments, encounters and a component to create an encounter for the undergoing visit.
• On creating the encounter a redux action updates active page to dashboard.
• the dashboard page contains 4 tab: home, vitals, allergies, medication.
• the home tab is the default active tab and contains already exisiting vitals, allergies and medications of the patient.
• the other tabs are used to create specific resource entry.
• the user can log out and select a different patient to repeat the above points.

Some of the required components are not merged in the main repo yet, hence respective tab is unimplemented.

Each worklow in the application is a different page with a unique url. Hence the use of redux in immense in the project.
Each workflow also has a dedicated reducer which handles all the operations related to a specific workflow without editing any other state.

Inside the project’s src directory, the workflow directory contains all the code distributed into separate directories for different workflows. This makes it very simple to create / edit or even delete a workflow without affecting other parts of the application!

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

import React from 'react'
import { Navbar } from '../../components/navbar'
import { Home } from '../../pages/home'
import { activePageName } from '../../reducers/visit/visitNurseSlice'
import { CreateEncounter } from './pages/createEncounter'
import { SelectPatient } from './pages/selectPatient'
import { Dashboard } from './pages/dashboard'
import { useSelector } from 'react-redux'

export const VisitNurse = () => {
  const activePage = useSelector(activePageName)
  let components
  switch (activePage) {
    case 'SelectPatient':
      components = [Home, SelectPatient]
      break
    case 'CreateEncounter':
      components = [Home, CreateEncounter]
      break
    case 'Dashboard':
      components = [Home, Dashboard]
      break
  }
  return (
    <Navbar labels={['HOME', 'VISIT']} component={components} />
  )
}

This is the main component that is exported. Based on different redux state values different screens are displayed.

All the implemented screens and the workflow can be seen here:

https://gitlab.com/librehealth/toolkit/lh-toolkit-webcomponents/uploads/a9e04307d1199767e3b2c400de785370/screen-capture__2_.webm

NEXT WEEK’s PLAN:

• Complete the visit-nurse workflow (the parts which are not implemented yet.)
• Complete the visit-provider workflow.
• Start working on the e-prescription workflow.

For the visit-nurse workflow, the medication statement and the vitals tabs are left to be implemented, these must be completed within the next week.

There are also some additions to be made in the webcomponents : some fields like encounter and subject (patient) are required to be added to the allergy and medication statement components.

The visit-provider workflow has similarities to the visit-nurse workflow, some of the screen are similar for both, hence the provider workflow should be completed faster than the nurse workflow.

these are the implementation points for the visit-provider workflow:

• First screen contains a find-patient component to select a patient.
• After the patient is selected a redux action is triggered which updates redux state and stores the selected patient Id.
• Another redux action updates the active screen and changes it to that of patient encounters.
• The patient encounter screens contains all the patient encounters which are fetched using the stored patient Id.
• On clicking an encounter a redux action updates active page to dashboard.
• the dashboard page contains 2 tab: physical exams, order.
• the physical exams tab shows the observations of the patient and contains a component to create any observation.
• the orders tab will be used to create orders using service request component.
• the user can log out and select a different patient to repeat the above points.

see you next week 😄.

WEEK5

PROGRESS MADE THIS WEEK:

Hi! We have a small change in the implementation plan. As per the proposal this week => week 5 was meant for implementing the visit workflow in the react / EHR application. The implementation required some new components, which were created last week as per the plan, but the Merge Requests are not merged yet. As there can be multiple changes and modifications required after the MRs are reviewed, It would be better to wait till the components are merged after modifications if required.

Therefore this week was based on doing the tasks of the 7th Week as Week6 is planned for improvements and bug fixing.

Primarily this week was spent creating the last set of components that will be required for the workflows planned.

The following components were created -

• MedicationRequest
• dosage

These components will be used for e-prescription workflow, although the dosage backbone element can be used for other Medication workflow based resources as well.

The pattern used for creating these components is same as the one used for other components.

Here is the screenshot:

next week:

• find and fix bugs in the react app.
• start implementing visit workflows in the react application.

see you next week 😄.

WEEK5

The main focus for this week was correcting the errors in the bounding box drawn. The bounding boxes differed from the ground truth. That was because of issue regarding the canvas context drawing setting and not related to the actual model.

After the correcting the drawing steps in the canvas context. The bounding box are drawn somewhat similar to the ground truth.

Second goal for this week was development of a remote server based object detection video api. I successfully mad it work for single image but beyond it the server started to hang-up.

The video encoding and decoding process are particularly stressful for the CPU. Throughout the development time the CPU was under 96~97% utilisation running at maximum clock speed. This was the case for a single instance, but if multiple instances are launched the server crashes instantaneously. For running the frozen inference model GPU was used. Under single gpu setup the gpu was under 100% memory utilisation.

Apart from this network latency spikes up during the encoding and decoding process.

That's all for today!!!!

Hope you had a great week

Adios

Settings drawer to toggle settings with unit test

• Created end drawer to toggle settings
• User can toggle notifications, notifications sound, dark mode
• The settings done by user are saved instantly in storage and changes are made according to it
• Wrote unit test for settings bloc with test cases 1. yields SettingsState on toggle of dark mode 2. yields SettingsState on change of notifications sound 3. yields SettingsState on toggle of notifications

Stage 3 Assessments Capture functionality

• Stage 3 has 3 types (normal, problem, danger)
• If a baby is classified as any type then that type of stage 3 assessments get added in app

Video GIF showing stage - 2 with classification functionality

• background turns red on danger
• background turns yellow on problem
• background turns green on normal

Stage 2 Unit testing

Test Cases for stage-2

• Throws exception if assesments are done before 60 mins from birth
• Throws exception if wardname is empty
• Throws exception if assessments are incomplete
• Marks phase-2 as complete if everything is correctly filled
• Adds stage-2 assessments if list is empty

Test Cases for Classification Repository

• if temperature < 95.5 or > 99.5 then classifies as danger
• if severe jaundice or fast breathing then classifies as danger
• if weight < 2000 & poor feeding then classifies as problem
• if everything is normal then classifies as normal

Showing summary of 24 hours in home screen

To show summary of 24 hours, We need to use lastUpdatedDuration=1d (1 day) as parameter in API to fetch all the events happened within 24 hours.

This has 3 types

• Admitted Count - Check Status = ACTIVE & Enrollment Date yesterday
• Discharged - Check Status = COMPLETE
• High Risk - For each TEI (Baby) this needs to be counted by parameter classification

What progress have you made this week?

• Phase - 3, 4 assessments capturing functionality
• Showing summary in Home Screen
• Developing settings screen of app
• Unit Testing phase - 2

What do I plan to do next week?

• Adding On-Call Doctors Viewing Functionality in App
• Completing Notification screen (Risk Assessment, Monitoring Alerts Tab)
• Phase - 3,4 Unit testing

Hey Everyone, I am back with my weekly blog. Let’s have a look at work done in the 6th week of this amazing journey.

Reimplemented and Completed App Usage Tracking System

Guys actually as stated in the last blog I had completed the app usage tracking system but on discussing with the mentor We come up with a better approach and reimplemented the completed system. There are not very huge changes we just did some changes in the dhis2 storage part. Let’s have a look at how we are tracking the trainer app usage now -

• App usage is now stored in the user’s data store instead of the tracker program.
• For each page, a unique id is assigned with which the page usage is tracked in datastore and in local DB as well.
• There is a local DB support that will store the page usage and when the page usage exceeds the threshold limit for each usage then data will be sent to the dhis2 in the user’s datastore.
• When local app usage storage get synced with the dhis2 user’s data store then local DB will get resettled to store new data.

Related MR: Added System to track app usage (!196) · Merge requests

Resolved Issues: #36 #120

Passed Pipeline:  #338127000 

Modified mhbs-tracker app to send login credentials

Since we want the user to enter credentials only once for both the apps and trainer app to be launched from the tracker app. Therefore the idea is, tracker app will share the login credentials with the trainer app on each launch. I had saved the login credentials of the user in secured Shared Prefs and on launching intent of trainer app passed these credentials so that trainer app to process further.

Related MR: Passed user credentials to trainer app for login & Fixed Some Warnings · Merge requests

Resolved Issues: #16

Passed Pipeline:  #338443846

Removed Fabric usages

The fabric plugin is depreciated and we need to remove its instances from the app.

Committed here.

Other Works

• Added commits to the previously sent MR as asked by the mentor to make them able to merge. As for testing purposes, I had used the hardcoded credentials, Now I had removed them and fetched the password from secure storage.
• I am not clear with the login procedure of the trainer app. I had studied the codebase and understand it well and made changes in the trainer app accordingly to make the user able to get logged in to the trainer app from the tracker app.

Work Done this Week

1. Implemented New Idea suggested by the mentor and Completed the system to get trainer app usage over dhis2. 
2. Removed Fabric usages from the app as it is depreciated.
3. Modified mhbs-tracker app to send login credentials to trainer app for login.
4. Added commits to the last MR’s to make them able to merge as suggested by the mentor.
5. Completely understand the secure storage and login flow of the trainer app.

Planning of next Week

1. Completed all pending work so can focus on new issues. Will be solving more bugs.
2. Working on “Need to create a new program stage in dhis2” issue of tracker app.
3. Working on #4 issue.

I am not clear with these issue so will try to discuss these issue in the weekend so that can start work from Monday.

Important Links

1. Added System to track app usage (!196) · Merge requests
2. Resolved Issues: #36 #120
3. Passed Pipeline:  #338127000 
4. Passed user credentials to trainer app for login & Fixed Some Warnings · Merge requests 
5. Resolved Issues: #16
6. Passed Pipeline:  #338443846

 Week 6 Coding period GSOC 2021 LibreHealth-

Hey, Reader's we are in the 6th week of the coding period.

The 6th week started on 9th July. This week I started working on the Cost of care web application. 

What progress have you made this week-

In the 6th week of the coding period, I have done the following thing with the cost of care web application.

1-added overpass API.

2- Display overpass API result like hospital name, distance, and no of beds available same as flutter application.

This week I was busy with college work. That's why I did less work as compared to the previous week.

My plan for next week- I have planned to do the following task for week 7

1-Improve Home screen UI.

2-I will send 1 merge request to cost of care web app project. 

I started working with the cost of care web react application  I am enjoying working with this project.

see you next week. 

 Find Minimum in Rotated Sorted Array II-

Suppose an array of length n sorted in ascending order is rotated between 1 and n times. For example, the array nums = [0,1,4,4,5,6,7] might become:

• [4,5,6,7,0,1,4] if it was rotated 4 times.
• [0,1,4,4,5,6,7] if it was rotated 7 times.

Notice that rotating an array [a[0], a[1], a[2], ..., a[n-1]] 1 time results in the array [a[n-1], a[0], a[1], a[2], ..., a[n-2]].

Given the sorted rotated array nums that may contain duplicates, return the minimum element of this array.

You must decrease the overall operation steps as much as possible.

 

Example 1:

Input: nums = [1,3,5]
Output: 1

Example 2:

Input: nums = [2,2,2,0,1]
Output: 0

 

Constraints:

• n == nums.length
• 1 <= n <= 5000
• -5000 <= nums[i] <= 5000
• nums is sorted and rotated between 1 and n times.

 

Follow up: This problem is similar to Find Minimum in Rotated Sorted Array, but nums may contain duplicates. Would this affect the runtime complexity? How and why?

 

question link -https://leetcode.com/problems/find-minimum-in-rotated-sorted-array-ii/

Solution-

To solve this problem we will use the binary search with some modifications. We observe that the minimum element is less than its left and right element.

we start searching start to end if start < mid element it mean our minimum element belong to another half.

code-

Hello Everyone, sorry for getting delayed in writing this blog. Last week's work is as follows: 

Worked on App Usage Tracking System of Trainer App 

To track the app usage I had followed these steps-

1. Created “mHBS Trainer Usage Tracker” Tracker Program

• Created Tracked Entity Attributes to store [ total page visits, total time spent on a page, page visits per user, time spent on a page by per user].
• Created “Page Usage Tracker” Tracked Entity with the above attributes.
• Created a tracker program i.e. “mHBS Trainer Usage Tracker” to track the tei created above (Page Usage Tracker).
• There are more than 40 pages in Trainer app. Enrolled tei for each page in “mHBS Trainer Usage Tracker” program.

Now we can send data for each and every page to dhis2 in the Tracker program.

2. Created Local SQL- DB table to store app usage

• Users cannot remain online every time but we want the full usage of our app to track therefore we need some local storage to store app usage in offline mode and update this data to dhis2 whenever the user comes online.
• Therefore created “app_usage” SQL table in the trainer app with the — [page_id (stores tei id of the page with which it is created at dhis2), page_name, page_visits, time spent on page ]
• Whenever a user visits any of the pages ( get to know from routes ) then we update its visits in local DB. 
• And similar to time spent on any page, it will be calculated with the help of “beforePageIn” & “beforePageOut” Events with simple subtraction of time in invoking this event.
• Then update this data to dhis2

3. Idea of Threshold Limit for app usage

• I thought that instead of hitting the API continuously to update dhis2 with current usage we can make a threshold limit like 50-page visits means whenever any page visits crossed the number 50 only in that cases we will hit the API to update dhis2 and in other cases, we just update the usage in local DB.
• This will help us to save a lot of users data.

All of the above-discussed flows are integrated and set up. But recently we come up with a better approach suggested by the mentor, so I will try to implement that one and send the MR instead of this approach.

Works Done in the Fourth Week

• Created “Page Usage Tracker” Tracked Entity & “mHBS Trainer Usage Tracker” Tracker Program
• There are approx 40 pages in the trainer app, made tei instance for every page in “mHBS Trainer Usage Tracker” program to keep its usage track.
• Stored page visits and time spent on each page by the user in local SQL-DB as well as updated this to dhis2 program. 

What do I plan to do next week?

• Complete Pending Tasks, reimplement this solution with a better approach suggested by the mentor.
• Try to fix some more bugs.

Have you had any blockers or issues that are impeding your project?

• Not yet

The main target for this week was POC development. I was successful able to develop a POC with help of HTML5 Canvas and WebXR. The total inference time is about 25 seconds on a Snapdragon 870.

The previous week's model had inherent Frozen-Graph Operations issue because of which is did not perform well when used with a larger dataset.So, I had to re-train the model. The newer performed well.

For Bounding Box generation, I have HTML5 canvas where in I draw the rectangle with the help of context2d. Total time for bounding box drawing and model inference is about 25 seconds for 3 min clip.

The video clip tested contains the images which are part of the segmented images obtained from the kavsir dataset. For joining the images I have used a software called as clideo.

The POC works well on Android phones without any major performance issues.

That's all for today!!!

Hope you had a great week

Au Revoir

References

Clideo

Stage 2 Assessments Capture with Classification

• Created stage-2 assessments capture with clasifying baby’s health using program rules
• Health is classified in three categories (Normal, Problem, Danger)
• App takes all data entered by user in stage 2 & then evaluate program rules for classifying health
• After classifying health the color of assessments screen changes according to the category

Stage 2 Notifications showing feature

• Used awesome_notifications package for displaying notification
• Displaying scheduled notifications reminder if user has not completed assessments
• Displaying notification on stage assessments opening
• Created Offline Notification showing functionality
  1. Showing notification when baby is registered to complete phase -1 assessments
  2. If user forgets to do phase 1 assessment then reminding by notification again after 60 minutes
  3. Notification feature works offline using scheduled notification featur
  4. Package used for showing notification https://pub.dev/packages/awesome_notifications

Screenshots of notifications

Unit testing assessments bloc, client, repository, phase-1

Test Cases

• Flutter Assessment BloC test
  • Test Cases
  • yield AssessmentsInitial on successful fetching of data
  • yields AssessmentsError on unsuccessful fetching of data
  • yields AssessmentsAdded on successful adding of phase 1
  • yields AssessmentsError when ward name is empty while adding of phase 1 assessments
  • yields AssessmentsError when assessments are not completed while adding of phase 1 assessments
    
• AssessmentsRepository testing
  • Test Cases
  • Throws exception if wardname is empty
  • Throws exception if assessments are incomplete
  • Marks phase-1 as complete if everything is correctly filled
  • Adds stage-1 assessments if list is empty
    
• Assessment Client Test
  • Test Cases
  • returns response with 200 code on successful fetching of assessments
  • throws fetch data exception on no internet connection

Searching & sorting functionality on list of babies screen

Sort List of babies by

• Time (Birth Time)
• Status ( Classificiation - Normal, Problem, Danger ) (Danger & Problem have high priority)
• Location ( Ward Name)

Search List of babies by mother’s name

What progress have you made this week?

• Adding Search, Sort Functionality in List Of Babies screen
• Unit Testing Assessments Bloc, Client, Repository, Phase 1
• Created Phase-1 Notifications showing functionality
• Created Phase-2 Assessments Capture with notifications

What do I plan to do next week?

• Phase - 3, 4 assessments capturing functionality
• Showing summary in Home Screen
• Developing settings screen of app
• Unit Testing phase - 2 , 3 , 4

 Week 5 Coding period GSOC 2021 LibreHealth-

Hey, Reader's we are in the 5th week of the coding period.

The 5th week started on 2nd July. This week was exciting and I enjoyed this week's task very well.

What progress have you made this week-

In the 5th week of the coding period, I have done the following thing with the cost of care flutter application.

1-Added address field To download CDM screen.

2- search by address and hospital name to download CDM screen.

3- added text when no bookmarked item left.

4-UI modification.

I sent 1 merge request to the development branch including all changes.

My plan for next week- I have planned to do the following task for week 6

1-I will start working on cost of care react web application. 

2-I will read the codebase of the web app.

3- I will display nearby hospitals overpass API.

I really enjoyed the 5th week of the coding period and I gained a lot of knowledge from it.

Now we are going to enter in  First evaluation period  (12-16 July ). 
see you next week. 

Stage 1 Assessments Capture Functionality

• Fetching & Capturing Stage - 1 Assessments , with this feature user is able to view the captured assessments & also capture the assessments and push the data to DHIS2.
• This works offline also, data is sent to server when user refreshes on having internet next time
• Created a BLoC to handle assessments
• Created Stage - 1 model class which parses and converts to JSON format for parsing and converting data
• added required ID’s in static variables in dhis2_config.dart file
• created a client to send data, created a repository to handle data from client

Fetching list of babies enrolled in program

• Created a BLoC for fetching list of babies enrolled in a program
• Creating a client & repository class for fetching & saving data locally
• Sorting the list by time to categorize recently & past registered babies
• On Tap of baby list item the app opens assessments screen & start fetching assessments
• List item color changes with classification of baby - Danger (Red), Problem (Yellow), Normal (Green)

Merge Request Link

User Profile Screen Functionality

• Fetching list of users notification from dhis2
• saving data in hive storage after fetching item
• fetching & displaying the notification date & time for each notification

Merge Request Link

Localized strings in assessments phases UI

• All strings used in assessments phases UI are now localized
• localized strings are saved in app_en.arb file

Merge Request Link

What progress have you made this week?

• Implemented User Profile Screen functionality
• Implemented fetching list of babies from dhis2 & displaying in screen
• Cache the data after fetching from dhis2
• localized strings in assessments phase
• Phase 1 assessments capturing functionality

What do I plan to do next week?

• Phase - 2, 3, 4 assessments capturing functionality
• Adding Search, Sort Functionality in List Of Babies screen
• Adding sort functionality , settings UI, activity logs in profile screen
• Update functionality in profile screen

WEEK4

PROGRESS MADE THIS WEEK:

Advancing into the 4th week, everything is going as per the plan. By the end of this week, most of the components planned are already created.

This week (4th week) was utilised creating components for the visit workflow.

The visit workflow primarily involves the nurse and the provider along with the patient, the nurse checks the arrival of the patient, takes the vitals, verifies and creates medications and allergies.

Provider then checks the patient and performs exams and orders lab tests if necessary.

FHIR does not have a visit resource so every operation related to a visit event is handled by the modified encounter resource.

In FHIR, Appointment is used for establishing a date for the encounter,

When the patient arrives and the visit is about to start, then the appointment will be marked as fulfilled, and linked to the newly created encounter.

The following components were created for this: (other components which will be required apart from the following mentioned components are already created)

• Encounter
• Service Request
• Medication Statement
• Quantity datatype component
• Allergy intolerance
• Observation

The web component repo already had components for allergy and observation, but their utility was limited, moreover for every property there was a components which was actually not required as using created datatype components all the properties of any resource could be included in a single component.

A definte pattern is used for creating these components. The pattern is demonstrated below:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37

/*imports required*/
import { LitElement, html, css } from 'lit-element';

class MyCustomElement extends LitElement {

  //all the properties are defined here
  static get properties() {
    return {
      value: { type: Object, reflect: true },
    }
  }

  constructor() {
    super();
    this.value = {}; 
  }
   
   // handle the change in input
   setDisplayValue(e) {
    this.value.display = e.target.value;
  }  

  // templates to render the component
  displayTemplate() {
    this.value.display = this.value.display || "";
    return html`<div>HELLO!</div>`
  }

  // render al the templates created 
  render() {
    return html`
      ${this.displayTemplate()}
    `
  }
}

customElements.define('my-custom-element', MyCustomElement);

Here are the screenshots:

—

—

next week:

• create the remaining component.
• implement visit workflows in the react application.

see you next week 😄.

WEEK4

This week was an eventful one. I started the week with TF Object Detection model development. I trained two models one was Faster RCNN and another one was MobileNet SSD. I plan to train YOLO models too and convert them into ONNX format for JS inference.

For VR the bounding boxes posed a lot of problems . For rendering the bounding boxes I used HTML5 Canvas and WebGL for designing and rendering the bounding box. The Dataset for Segmentation has only one class so class labels are not rendered on top of the bounding box. The model inference is still slow and in some cases fails to render the bounding boxes.

POC for Object detection is not working properly at this juncture. It still requires few more corrections for improving the model performance.

That's all for today!!!

Hope you had a great week

Vale

Hello everyone, do you want to know what is the progress of our project?😏

Obviously, I am here to share that with you. So let’s have a look at the work done in the fourth week of this amazing journey🧐-

This week I had set up Gitlab CI for our trainer app, Completed the offline support for the trainer app using SQLite database, added syncing UI in the app to differentiate the downloaded content, testing 2–3 issues that were mentioned in the last blog that they are resolved or not. Wait guys let’s discuss all this in detail😀-

Without CI/CD it difficult to check whether the particular code will build properly or not and we have to think about a local clone of the project or any other option to get its latest apk. To solve this I had set up the CI/CD Pipeline as well with the MR sent last week and in the one sent this week also.

Resolved Issues: #112

committed here - ce83405a 

Last run Pipeline: #327895681

CI is also added to offline-support MR.

Initially, in the last MR documents on the media page and every media file get downloaded and then played there is not any support for storing any type of data locally to the device to make it available offline.

Therefore I had added an SQL database to the trainer app to store data. This week I had completed my pending work related to this which was just started last week.

Now the updated list of documents get saved to the local database whenever a user visits the media page with an active internet connection and if a user visits this page in offline mode even in that case he is able to use the application in the same manner as it is in online mode, in this case, it is fetching data from a local database.

When a user clicks on any media file if it does not download it takes some time to get downloaded the first time and show a progress UI. Let’s suppose if it is already downloaded in that case it gets played instantly without waiting for a newer version but every time when user clicks on the media either it is already downloaded or not, it hits the API to update the media file. In offline mode, nothing update happens and it played the last cached version stored in the device and in online mode, it instantly plays the last cached version and hit the API to listen for any updates on this file.

SQL Database Schema & Logic

1. When the application launches it creates a media table with 4 columns-

• id- stores the id of the document with which it is stored on dhis2
• name- stores display Name of the document
• fileuri- will be not null only when the document is downloaded and saved on the device, it stores the local location of the document on the device and in other cases its value is null
• mimeType- it stores the content Type or media type of the document like application/pdf, video/mp4, etc. to play the media using the fileopener2 plugin

2. Logic for storing a list of documents of media page-

• when a user visits this page new list will be fetched from dhis2 and now we need to take care of 3 cases : 

(1) DELETE : Document ids not present in the new list but present in our local DB. These documents are deleted from dhis2 and need to be deleted from local DB as well.

(2) UPDATE : After completing the first step, local DB is left only with those ids which need an update. We just update the particular row of their ids of local DB with whatever new corresponding data we received.

(3) CREATE : Now those document ids of a new list that are not passed through step 2 are one of those of newly added documents. We just need to create a new document with their data in our local DB.

These steps will take care of all cases for making a list of documents available offline.

committed here: e8b93759

3. Logic for offline support of media files-

• Whenever the user clicks on the media file it gets downloaded and saved to the device local storage and its corresponding file location and mimeType gets updated to the fileuri and mimeType key of local DB for that document. And on the next time, a user doesn’t need to wait for it to get downloaded it gets played from the last saved location.
• I had used PERSISTENT storage and cordova-file-plugin to save the file to the device.

committed here: 65ddad02

Resolved Issues: #69

Corresponding MR: !195

Pipeline: #329574660

I had added 3 types of sync symbols ( Pending, Completed, Error ) in the media page list.

I had also added a progress bar that shows the ongoing sync state.

committed here: bf49a629

There are some issues whose testing I have to do last week. Let’s discuss them one by one-

1. Track app usage data ( #36 )

• We want to track the usage of our app and gets its report on dhis2 or tracker app.
• For that, the page visit is set up for all pages means whenever the user visits any page then it gets recorded on local device storage as a counter of visits to that page.
• Now the need is to shift this data to dhis2 via Web API in an App Usage Tracker Program. This part is not done.

2. Role-based restricted access for viewing tracked entities ( #44 )

• We want that the TEI should be visible only to the person who had created it other person visibility to this should be restricted.
• This part is not working and since we had updated the mHBS app from dhis2 after 3 years it has a lot of changes and the new codebase is completely in Kotlin and well architectured and segregated.
• Therefore this part need rework. This is not working in the new app. You can have a look at the latest mHBS tracker app codebase from here.

3. Need to create a new program stage in dhis2 ( #10 )

• This issue is also cannot be closed for now.
• There is no commit in the latest code that can solve this for now.
• Needs rework.

Works Done in the Fourth Week

1. Set up CI/CD to automatically build the apk and check for errors.
2. Added full offline support for the media page for both media files and a list of documents.
3. Added file synced based UI for each media file.
4. Tested 3 Issues.

Important Links

1. Merge Request: Having code related to full offline support & SQL implementation of media page.
2. Solved Issue: #69 #112
3. Gitlab Pipeline : #329574660 #327895681
4. Tested Issues : #36 #44 #10

Work to be done in the next week

1. Working on Issue #36 

• Shifting all app usage data from local storage to dhis2

      2. Working on Issue #44

• Restricting the access of other users for viewing tracked entities

      3. Working on Issue #10

• create a new program stage in dhis2

 Week 4 Coding period GSOC 2021 LibreHealth-

Hey, Reader's we are in the 4th week of the coding period.

The 4thweek started on25th June. Till now I am really enjoying this coding period and I am improving my development skill.

What progress have you made this week-

In the 4th week of the coding period, I have done the following thing with the cost of care flutter application.

1- Added Search bar to compare hospital screen.

2-Compare Hospital Screen UI modified. 

I sent 1 merge request to the development branch including all changes.

My plan for next week- I have planned to do the following task for week 5 

1-I will add the Search by address feature to the download CDM screen.

2-I will add a new address field in download CDM  list.

3- I will discuss UI regarding the new address field with mentors.

The 4th-week tasks were quite easy. During the 4th  week, I sent a Merge request to the development branch as I told you earlier. 

This is all about 4th week of the coding period. I really enjoyed it.

See you next week.

WEEK 3:

PROGRESS MADE THIS WEEK

Hello everyone. As dicussed in the previous blog, this week was to be utilised for creating the Appointment workflow in the React application.
Surpisingly, the task hardly took 24hrs for completion, henceforth by the end of Tuesday I was done with this week`s work 😄.
Anyways there were a few bugs in test files that were to be fixed, so I utilised my weeks fixing some web-component tests.

here are the screenshots of the workflow:

upcoming days:

• start working on the task of the coming weeks.

next week:

• complete the components for the visit workflow.

see you next week 😄

WEEK 3:

WEEK 1:

Finally the community bonding priod is over and the coding period began. In this blog I have loads of content and code to cover!

ANALYSIS OF COMMUNITY BONDING PERIOD
In the community bonding period there were mainly 2 objectives we had set :

• SETUP THE REACT APPLICATION WITH CI/CD.
• DESIGN THE APPLICATIONS WITH THE ENDPOINTS.

Most of my community bonding period was spend on the CI/CD part. The major cause of this was
cross-browser cross-platform testing integration with React, the tool which was planned to be used was saucelabs and unfortunately for me :(, there was no proper content regarding saucelabs integration with react to be found anywhere, I tried to make things work on my own and nearly succeeded testing on jest with snowpack and react, but snowpack could not resolve the dependencies inside the lh-toolkit-webcomponent monorepo when it was caching the files.

Snowpack serves the application unbundled during development. Each file needs to be -built only once and then is cached forever. When a file changes, Snowpack rebuilds that single file

As mainly we will be covering unit and integration testing, saucelabs will not be a neccessity as suggested by the saucelabs community. So I went on to setup the react app

• testing : using react-testing-library with jest,
• deployment : gitlab pages
• some eslint integration and git hooks.

Design part was smooth sailing and offered many insights that I might have not noticed until later, upon the suggestion of my mentor I have also documented the api endpoints. I will drop the file with the designs and the endpoints here : https://geforce6t.github.io/blog/categories/EHR/

PROGRESS MADE IN THE WEEK

the plan for the first coding week is to do the following:

• setup the application
• create home, login pages
• implement the checkin workflow

monday: I started coding 🎉️ , and completed the first part of the plan.

tuesday: I started working on the project layout and accomplished the following :

• organize project into different directories including pages, components, redux store, redux slices etc
• added MUI with theme
• created login and auth pages with support for other workflows in mind.
• created redux auth workflow

wednesday: looked for ways to style the lit-components………, lit components uses shadow dom and the style encaptulation feature make it a little tricky to style custom components from the outside, I made some hacks to do the same using callback refs inside the react component where the lit component resides,but this approach would not be optimal.

Probably some minimal styles could be applied inside the lit components and would do great, anyhow I was able to style the material components using the global material css variables.

One of the things that I need to look into is to add some style to the lit-components as a whole : “margin, broder shadow and stuff”, this will not be a big issue as I have a lot of other components to create so mainly the style will be global and can be done anytime.

So here is how the create-patient component has changed visually:

compared to above, this is what we have as default:

thursday: created the general search component.

the general serach component is something that I would like to talk about. The general search component was actually little difficult thing to do, for different cases the search params, columns, query parameter may differ so to create a general purpose search component all of this features must be added as an attribute which was done in the implementation,

the functions used in the seach component are

• one for search, based on fetching and storing the result after filtering the result to only include results for the columns using the query.
• another for storing pagination contexts: next page and previous page url,

the most difficult part about the implementation was filtering results to form column enteries, the values that we may want can be nested one/many levels so it is very important to provide a query string to execute this filtering, it is not possible to pass this query string via props as it is inside a map method, so I created a function with switch case for different workflows that we will implement.

finally the components looks like this :

friday: wrote the edit and show patient code to complete the layout of the checkin worklfow 🎉️, but there was one issue that came across while creating the put request, for some of the fhir components the values are not changing even after visually their value is changed, this might be an issue in the react app or the wc repo. I will take a look and solve this issue in the coming days!

UPCOMING DAYS

• add event handlers for the create patient page.
• debug the value not changing issue while making a put request.
• add css for element (not compulsary for now)

thanks for reading till the end ❤️ , see you next week 😄

WEEK 1:

Finally the community bonding priod is over and the coding period began. In this blog I have loads of

WEEK 2:

Hello folks! Another week done and another batch of code written and merged !! In the coming sections I will be explaining the details of the work done this week.

PROGRESS MADE THIS WEEK

In the previous week, I had implemented the checkin workflow in the react application. According to my plan in the proposal, this week was to be used to create components for the resources that will be used in the Appointment workflow.

the resources are:

• Slot
• Schedule
• Appointment

In the beginning of the week, I started looking into the lit project: upon which the components are based. I read the documentation and decided how to create the components so that they can be utilised for different CRUD operations.

I also noticed a few bugs in the components while implementing the checkin workflow, for this I created a MR and fixed them.

By wednesday I started creating the components. Most of the components that will be created have a reference or codeable-concept property datatype, hence creation of components that can be used in different contexts to implement the mentioned datatypes will help a lot.

By the end of wednesday I had created these datatype components. Once these components were created, the resource components were not difficult to develop.

By thursday I had created Slot-Schedule and by friday Appointment.

During this period my mentor noticed a few bugs in the project:

• inconsistent label case use
• unavailable demo files (build issue)

So along with the development of the components, I made a couple of commits for these issues as well.

Template used for component creation:

As a component will be used for CREATE / GET / PUT request , I made it compatible with everything.

For the create part, inside the constructor an initial empty skeleton value was provided. This value property would be overwritten if the component contains the value attribute. Hence this supports both the cases of create and get operations as a user can avoid the value attribute if the intention is to use the component for creating entries and provide a value after fetching a specific endpoint if the intent is to show entries.

Moreover even if a value property is provided and some of the keys are missing, the component will still remain the same as I have done a check for individual property existence and default value allocation as well, this feature will be helpful for the edit operation.

here is the default value property set inside the Appointment component constructor. Similar structure is used in the other 2 components.

1
2
3
4
5
6
7
8
9
10
11
12
13
14

{
      identifier: [{}],
      status: '', 
      serviceCategory: [{coding: [{ system: "", code: "", display: ""}],text: ""}],
      serviceType: [{coding: [{ system: "", code: "", display: ""}],text: ""}], 
      specialty: [{coding: [{ system: "", code: "", display: ""}],text: ""}],
      appointmentType: {coding: [{ system: "", code: "", display: ""}],text: ""},
      slot: [{reference: "", display: "", type: ""}],
      comment: '',
      start: moment().format('YYYY-MM-DDThh:mm:ss[Z]'),
      end: moment().format('YYYY-MM-DDThh:mm:ss[Z]'),
      participant: [{type: [{coding: [{ system: "", code: "", display: ""}],text: ""}], status: '', actor: {reference: "", display: "", type: ""}}]
    }

here are the screenshots of the implemented components:

upcoming days:

• write unit test for the components.

thanks for reading till the end ❤️ , see you next week 😄

WEEK 2:

Hello folks! Another week done and another batch of code written and merged !! In the coming sections I

CHECK-IN WORKFLOW :

Valid search parameters for this search are: [_id, _language, _lastUpdated, active, address, address-city, address-country, address-postalcode, address-state, address-use, birthdate, death-date, deceased, email, family, gender, general-practitioner, given, identifier, language, link, name, organization, phone, phonetic, telecom]`

SEARCH PARAMS :

GET [base]/Patient?identifier=[value]

GET [base]/Patient?name=[value]

CREATE PATIENT :

POST [base]/Patient

GET PATIENT :

GET [base]/Patient/[_id]

EDIT PATIENT :

PUT [base]/Patient/[_id]

check-in.mp4 from Shashwat on Vimeo.

APPOINTMENT WORKFLOW :

SCHEDULE TAB

Valid search parameters for this search are: [_id, _language, _lastUpdated, active, actor, date, identifier, service-category, service-type, specialty]`

SEARCH PARAMS :

GET [base]/Schedule?actor=Practitioner/[value]
GET [base]/Schedule?actor=Location/[value]
GET [base]/Schedule?actor=Patient/[value]

GET [base]/Schedule?service-category=[value]

CREATE SCHEDULE :

POST [base]/Schedule

GET SCHEDULE :

GET [base]/Schedule/[_id]

EDIT SCHEDULE :

PUT [base]/Schedule/[_id]

SLOT TAB

Valid search parameters for this search are: [_id, _language, _lastUpdated, appointment-type, identifier, schedule, service-category, service-type, specialty, start, status]`

SEARCH PARAMS :

GET [base]/Slot?status=[value]
GET [base]/Slot?schedule=[value]

CREATE SLOT :

POST [base]/Slot

GET SLOT :

GET [base]/Slot/[_id]

EDIT SLOT :

PUT [base]/Slot/[_id]

APPOINTMENT TAB

Valid search parameters for this search are: [_id, _language, _lastUpdated, actor, appointment-type, based-on, date, identifier, location, part-status, patient, practitioner, reason-code, reason-reference, service-category, service-type, slot, specialty, status, supporting-info]”`

SEARCH PARAMS :

GET [base]/Appointment?actor=[value]
GET [base]/Appointment?status=[value]

CREATE APPOINTMENT :

POST [base]/Appointment

GET APPOINTMENT :

GET [base]/Appointment/[_id]

EDIT APPOINTMENT :

PUT [base]/Appointment/[_id] `

VISIT WORKFLOW:

///////////////////////////////////////////////NURSE///////////////////////////////////////////////

FIND PATIENT:

SEARCH PARAMS :

GET [base]/Patient?identifier=[value]
GET [base]/Patient?name=[value]

once the nurse selects a specific patient the patient Id will be
stored by a redux action. So the patient id below will be fetched from
the redux store via a useselector hook to get the appointments and the
encounters related to a specifc Patient entry.

APPOINTMENTS RELATED TO THE PATIENT:

GET [base]/Appointment?actor=Patient/[_id]

ENCOUNTER RELATED TO THE PATIENT:

GET [base]/Encounter?subject=Patient/[_id] /* TO GET ALL THE ENCOUNTERS SPECIFIC TO A PATIENT */

POST [base]/Encounter /* TO CREATE AN ENCOUNTER */

PUT [base]/Encounter/[_id]

ALLERGY RELATED TO THE PATIENT:

GET [base]/AllergyIntolerance?patient=Patient/[_id]

POST [base]/AllergyIntolerance /* TO CREATE AN ALLERGY ENTRY FOR THE PATIENT */

MEDICATIONS THE PATIENT IS TAKING/HAS TAKEN:

the fhir resource used here is MedicationStatement : this resource
has a reference to the patient, this resource is required to know what medications the
patient has been taking / taken before.

GET [base]/MedicationStatement?subject=Patient/[_id]

POST [base]/MedicationStatement /* TO CREATE A MEDICATION-STATEMENT ENTRY FOR THE PATIENT */

PATIENT VITALS:

IN FHIR THERE IS NO SEPARATE RESOURCE FOR VITALS , VITALS CAN BE TAKEN
AS AN ENTRY OF OBSERVATION RESOURCE UNDER THE CATEGORY OF VITAL-SIGNS

GET [base]/Observation?category=vital-signs&subject=Patient/[_id]

POST [base]/Observation /* TO CREATE AN OBSERVATION ENTRY FOR THE PATIENT VITALS */

PUT [base]/Observation/[_id] /* TO EDIT OBSERVATION ENTRY FOR THE PATIENT VITALS */

/////////////////////////////////////////PROVIDER/////////////////////////////////////////

FIND PATIENT:

SEARCH PARAMS :

GET [base]/Patient?identifier=[value]
GET [base]/Patient?name=[value]

once the provider selects a specific patient the patient Id will be
stored by a redux action. So the patient id below will be fetched from
the redux store via a useselector hook to get the
encounters related to a specifc Patient entry.

ENCOUNTER RELATED TO THE PATIENT:

GET [base]/Encounter?subject=Patient/[_id] /* TO GET ALL THE ENCOUNTERS SPECIFIC TO A PATIENT */

PUT [base]/Encounter/[_id]

similar to the above case once an encounter is selected its id will be stored via redux store

OBSERVATION RELATED TO THE PATIENT IN THE CURRENT ENCOUNTER:

GET [base]/Observation?encounter=Encounter/[_id]

POST [base]/Observation

ORDERS RELATED TO THE PATIENT IN THE CURRENT ENCOUNTER:*

GET [base]/ServiceRequest?encounter=Encounter/[_id]

POST [base]/ServiceRequest

E-PRESCRIPTION WORKFLOW:

////////////////////////////////////////////PRESCRIBER//////////////////////////////////////////

FIND PATIENT:

SEARCH PARAMS :

GET [base]/Patient?identifier=[value]
GET [base]/Patient?name=[value]

once the nurse selects a specific patient the patient Id will be
stored by a redux action. So the patient id below will be fetched from
the redux store via a useselector hook to get the
encounters related to a specifc Patient entry.

ENCOUNTER RELATED TO THE PATIENT:

GET [base]/Encounter?subject=Patient/[_id] /* TO GET ALL THE ENCOUNTERS SPECIFIC TO A PATIENT */

PUT [base]/Encounter/[_id]

similar to the above case once an encounter is selected its id will be stored via redux store

ALLERGY RELATED TO THE PATIENT:

GET [base]/AllergyIntolerance?patient=Patient/[_id]

MEDICATIONS THE PATIENT IS TAKING/HAS TAKEN:

the fhir resource used here is MedicationStatement : this resource
has a reference to the patient, this resource is required to know what medications the
patient has been taking / taken before.

GET [base]/MedicationStatement?subject=Patient/[_id]

MEDICATIONS PRESCRIPTION:

the fhir resource used here is MedicationRequest

POST [base]/MedicationRequest

CHECK-IN WORKFLOW :

Valid search parameters for this search are: [_id, _language, _lastUpdated, active,

This was the third week of the coding phase and it was good. This week majorly, I had worked on the way to play the video blobs that I had fetched from the server last week and showing the pdf files in the app. I had worked on the offline support of the application as well. Let’s see everything in detail-

Added a feature to play all type media files in app

We have 2 types of media files videos and pdfs which we have to show in our app’s media section. Let’s see what approach I had chosen to achieve that and why?

1. Firstly I had downloaded the blob object of the media that is being uploaded on the dhis2 using XMLHttpRequest.
2. Then I had created a local temporary space in the memory to save the file.
3. And lastly using the fileOpener2 plugin, I had opened that file using the supported apps that are already available on the user’s device.

This seems simpler but really It takes a lot of time to figure out how actually Cordova handles all this stuff🤔because I am not much familiar with the Cordova platform but finally I had made it. Actually, for videos I had initially used the most popular thing that is iframe but it didn’t work and I had changed it with the video element of HTML, and it gets working 😍. But the problem arises when I am trying to open pdf because in the browsers there are inbuilt plugins that handle pdf opening and all but for the android, there is no such thing therefore I need to use external apps with the help of fileOpener2 plugin. 

Final Output: Now one can upload pdfs and videos on the dhis2 resources and it will be fetched to our trainer app and can be accessed with the internet.

Have a look at the PR related to this task: Added mediaPage and Populated with real data Fetched from bmgfdev instance and Added support to play all media files of type blob (!194) 

Related Issues : #190 #77 #39

Added SQLite Database for storing data locally

We want our application to be active in offline mode as well therefore I had decided to use the sqlite database of the cordova plugin. Let’s look at the implementation part of the offline-support of media section:

 1. I had made a mhbsTarinerDb and media table to store the data of all the cuments. I will be storing a name (displayName of the document), id (dhis2 id of the document on the server), fileUri (location of the file if this media is downloaded). 

2. When the users click on the media tab then a list will be made from the documents available on the local db and it is shown on the tab at the same time an API call is initiated to the server if a user is online and we received a success response we will update our local db and media tab with this new data.

3. When a user clicks on particular media files then as stated earlier a blob object is downloaded and from it, we make a local file and save it to the device, now one point gets added that is we will store its location to our db as well and when the user again clicks on the particular files then similar thing happen means old content is shown till the new one is not downloaded and when new content is ready it gets updated to our db.

This part is almost ready and is under testing you can look at its progress from the offline-support branch of my forked repo. Worked done till now in offline support of media tab is here.

Work Done this Week

1. Added a feature to play all type of media files in the app.

• The list of all documents uploaded on the bmgfdev instance ( here ) is visible in the app.
• Videos Media & Pdf files both can be opened and can be accessed through the app.

      2. Added SQLite database to store data locally and to provide offline support. 

Planning of next Week

1. Finish pending work of Offline support and create PR for that.
2. Adding CI to Trainer App: Continuous integration through Gitlab CI (#112) 
3. Testing of some issues — 

• Role based restricted access for viewing tracked entities (#44) 
• Track app usage data (#36) 
• Need to create a new program stage in dhis2 (#10)

Important Links

1. Added mediaPage and Populated with real data Fetched from bmgfdev instance and Added support to play all media files of type blob (!194)
2. Issue #190
3. Issue #77 
4. Issue #39
5. Added SQL Database for storing data locally (26ad4006) · Commits · Bhavesh Sharma / mHBS-trainer · GitLab

Basic POC (Proof of Concept) design was completed and successfully implemented using WebXR and react. A JSON file was created which stores the links and names of all image file present in the directory.

The BiT model was converted using TensorflowJS model converter. The model then generated was used in POC. Images are cast on a Curved Plane in the VR environment, on which model inference runs.

Every 5 Seconds a new image is loaded and inference is given with 4 seconds. The labels are displayed as Text on top of the Plane in the VR environment.

The Video demonstrates a sample scenario. The inference reports undefined because the model was not able to classify the successfully

That's all for today!!!

Hope you had a great week

Do svidaniya

Coding Period Week 3 GSOC 2021 LibreHealth-

Hey, Reader's we are in 3rd week of the coding period.

The 3rd week started on 18th June.

What progress have you made this week-

In the 3rd week of the coding period, I have done the following thing with the cost of care flutter application.

1- Implement Unit testing for inpatient procedure.

2-Implemented Unit test for outpatient procedure.

My plan for next week-

The 3rd-week tasks were quite easy. During the 3rd  week I sent a Merge request to the development branch. 

This is all about 3rd week of the coding period. I really enjoyed it.

See you next week.

The previous model trained was not performing well on the the testing dataset. So, I had to try a new approach for increasing the model accuracy as well as other parameters. When I tried training with other pre trained networks, after one epoch model started to overfit. So, I had to get creative.

I Used a  BiT model for training, only the last layer was trainable. The model was able to perform better than pervious pre-trained model based on MobileNet.

Regarding the VR Part. The WebXR setup was successful. I was able to set up a basic WebXR environment. TensorflowJs was used for model inference. I used react-framework for binding TensorflowJS and WebXR together. For access to onboard device sensors the WebXR content has to be served through HTTPS.

Model Training Results

MobileNet Based Model: loss: 0.9046 - acc: 0.7174 - f1_m: 0.2450 - precision_m: 301254.4277 - recall_m: 0.2275 - val_loss: 1.0795 - val_acc: 0.6811 - val_f1_m: 0.2366 - val_precision_m: 248592.8750 - val_recall_m: 0.2270

BiT Based Model:loss: 0.6331 - acc: 0.8733 - f1_m: 0.8899 - precision_m: 195757.4552 - recall_m: 0.9773 - val_loss: 1.2562 - val_acc: 0.7955 - val_f1_m: 0.9079 - val_precision_m: 126641.6953 - val_recall_m: 1.0619

Inference Time:

When the SavedModel is converted into TensorflowJS model using the converter few seconds get added up in the inference time. So far, the inference time is not that much and is close if not less than 1 sec.

That's all for day!!

Hope you Had great Week

Ciao

Coding Period Week 2 GSOC 2021 LibreHealth-

Hey, Reader's we are in 2nd week of the coding period.

The second week started on 11th June.

What progress have you made this week-

In the second week of the coding period, I have done the following thing with the cost of care flutter application.

1-Added New bookmark feature 

2-add to Bookmark chargemaster.

3-remove to bookmark chargemaster.

4-chargemaster UI modification.

5- added snacbar while adding or removing form bookmark.

My plan for next week-

During 2nd week I added another commit to the development branch and I updated the Merge request. 

This is all about 2nd week of the coding period. I really enjoyed it.

See you next week.

Internationalized Flutter app by adding support for other languages

Created .arb (Application Resource Bundle) file which contains key-value pair strings, I localized all the string text using till now in app. For now I have added translations for 3 languages only

Added metadata of ECEB & on call doctors programs

DHIS2 provides a metadata repository containing metadata packages with various content. A metadata package is a DHIS2-compliant JSON document which describes a set of metadata objects. I exported this metadata for the ECEB & On call doctors program. Users who will use this app will need to import this metadata in their dhis2.

Unit testing for authentication screen

Unit tests are handy for verifying the behavior of a single function, method, or class. I created unit test for the auth functionality.

Test Cases for Hive Storage

1. User loggedIn return true
2. Not LoggedIn return false

Flutter BloC test cases

1. yield AuthenticationLoaded on successful login
2. yield AuthenticationError on unsuccessful login

Auth Client Test Cases

1. response code 200 - successful login
2. response code 401 - unauthorized (wrong credentials)
3. Socket Exception - no internet connection

Register Baby screen functionality

• Baby able to get registered using app & their data is stored using api in DHIS2.
• This feature works offline also, data is synced when user is online next time
• Created a refresh bloc for state management
• Added Snackbars for showing info related to syncing
• stored the list in hive storage

What progress I made this week?

• Internationalized Flutter app by adding support for other languages
• Added metadata of ECEB & on call doctors programs
• Unit testing for authentication screen
• Register Baby screen functionality (using DHIS2 API)

What do I plan to do next week?

• Implementing User Profile Screen functionality
• Baby Health Assessments Capture Functionality
• Fetching List Of Babies screen in App using API

Creating Login User functionality

• Used flutter Bloc architecture for state management
• Used DHIS2 Auth API for authentication
• username & password is passed as Basic Auth in header in base64 encoding in API
• Users can now login using username & password in App
• Added Dark mode support
• Created config files to store dhis2 apis & id
• Added password hide/show feature

Screen Shot

Created UI for Baby Assessments

Assessments are divided into multiple phases-

1. Phase 1 Assessments 0 - 90 minutes from birth

2. Phase 2 Assessments 60 - 90 minutes after phase 1

3. Phase 3 Assessments

Categorized in 3 parts

1. Normal
2. Problem
3. Danger

According to the Classification the appropriate assessment will be shown to user

Screen Shots

Added Theme for Dark Mode

Toggle feature is pending but I have enabled support for dark mode

Screen Shots

What progress I made this week?

• Completed authentication/login functionality
• Created config files for storing API, DHIS2 Id’s
• Developed UI of all three Assessments screens
• Added support for dark mode

What I have planned to do next week?

• Register Baby screen functionality (using DHIS2 API)
• Implementing User Profile Screen functionality (using DHIS2 API)
• Making the above features work also when user is offline and sending data when user is online next time

Hello Guys! The most exciting thing has begun that is the coding phase of GSoC-21. My community bonding period was gone good in discussing things and setting up the environment. This blog contains the work done by me in the First Week of the coding phase.

First of all, I want to share one important point with you :

1. We had shifted our mHBS-Tracker app to GitLab from Github. Its new link is here.

Merged latest changes of dhis2/dhis2-android-capture-app

Since our repo is very much lagging the new changes of the dhis2 app. The most important task that should be done at the first priority is to merge all the upstream changes. Our repo has approx. 1118 commits whereas in today’s time dhis2 app has approx. 8K commits. Therefore as a first task, I had merged all the upstream changes to our repo in order to work on the dhis2 app.

Modified the dhis2 app to work as mHBS app

After merging the upstream changes we are ready to work on the dhis2 app. Now to use it as a mHBS Tracker app we need to do some modifications to it as per our need. Therefore I had made all those UI changes and App Data changes to make it work as per the needs of the mHBS Tracker App.

List of modifications and other features :

1. App Name and Logo has been changed.
2. UI Improvements.
3. Added External others app of mHBS to app drawer.
4. Added functionality to open other mHBS apps.

Screenshots of the New mHBS App :

Setup CI/CD Pipeline to automatically build apk/artifact

Without CI/CD it difficult to check whether the particular code will build properly or not and we have to think about a local clone of the project or any other option to get its latest apk. To solve this I had set up the CI/CD Pipeline as well with the MR sent.

Fixed some of the known build errors

There are various build errors that were causing the problem in the successful build of the application I had solved a few of them. But still, there are some issues which need to be resolved but they are not interrupting the apk to build, these are just be reflected when apk is building. You can get a successful build and the latest apk from the code now as well.

Works Done in the First Week 

1. Merged the upstream changes that have more than 6K unmerged commits.
2. Modified the final merged code to have features and look like the mHBS app.
3. Set up CI/CD to automatically build the apk and check for errors.
4. Fixed Some of the build errors.

Important Links

1. Merge Request: Having code of all the work done in the first week. It has been merged to the main branch successfully.
2. Last Triggered CI Pipeline: This pipeline was triggered lastly in the MR.
3. Last Triggered Job and Artifact Link: Contains all the build log printed while running the ‘assembleDebug’ job. You can download the apk from the right panel of this job page from the ‘Job Artifact’ section.
4. CI/CD Setup Issue: Link of the CI/CD Setup issue solved with this MR.

Work to be done in the next week 

1. Solved the build errors of the mHBS Tracker app. There are mainly two problems for the mHBS app :

• Fix error ‘ > java.lang.NoClassDefFoundError: javax/annotation/Generated ‘ in gradle build when building with jdk 9+
• Fix Error in gradle build — string too large to encode using UTF-8 written instead as ‘STRING_TOO_LARGE’

2. Start working on the mHBS Trainer app for the media section.

3. Check for the complete smooth working of mHBS app and Improvements.

Coding Period week 1 GSOC 2021 LibreHealth-

Hey, Reader's we are in 1st week of the coding period.

The coding period started on 8th June. I started working with the Cost of care flutter Application it was really exciting.

What progress have you made this week-

In the first week of the coding period, I have done the following thing with the cost of care flutter application.

1-upgraded project to futter latest version 2.2.1.

2- dart migrate to null safety. 

3-Updated all dependencies that are used in the project to their latest version.

4-Replaced Deprecated Widget.

5-some of the methods were deprecated I updated them too. 

My plan for next week-

During 1st week I sent one  Merge request to the development Branch but I was struggling with pipeline failure. Then I asked my mentors they helped me to fix the pipeline. The pipeline was failing because some of the bloc test methods and Dio library methods were deprecated. 

This is all about 1st week of the coding period.I really enjoyed it.

See you next week.

Developing UI of Register Baby Screen

I worked on creating UI of register baby screen, I have used the following packages

• Flutter Customizable Slider for weight, temperature input (Link)

• Flutter Date Time Picker to pick baby’s birth date & time (Link)

• Flutter Bloc Architecture to handle state management (Link)

I have created toggle buttons for baby’s information input such as gender, mode of delivery etc. These toggle buttons uses boolean variable for their state management.

I have created a model class for storing Baby’s health information, This class’s object is created & data is filled when user enter’s data is register baby’s screen.

Creating Navigation Drawer in App

I Added a navigation drawer in app, It has the following items & header with LibreHealth’s Logo.

• Doctor’s Schedule
• Messaging
• About
• Share App
• Report Bug

Screen Shots

What progress I made this week?

• Developed UI of Register Baby Screen
• Created Sliders for weight, temperature etc.
• Created Model Class for storing Baby’s Information
• Created Bloc for register baby screen
• Created App Drawer with items ( Doctor’s Schedule, Messaging, Share App, Report Bug)

What I have planned to do next week?

• Developing Assessments Screen UI for all stages
• Using Sliver App Bar for floating app bars
• Adding Dark Mode in App

This week was all about model building and model validation. Initially, I was stuck a bit on the metric portion, but was later able to overcome it by using special keras libraries. The labyrinthine keras and tensorflow documentation proved to be invaluable.

Model Development

Dataset

The dataset has labelled-images, unlabelled-images, labelled videos and segmented images. For the classification process I am going to use the labelled-images. They have 22 classes labelled in them.

Model

I am training two model, out of which one model will be a custom model and other one will be pre-trained one which is adapted to Kvasir Dataset. A size of 224x224 will be used as the input shape for the model. For data logging purpose i have used tensorboard.

Results

Both the models were trained. The pre-trained model provides low generalisation error in-comparison to the Custom-Model.

CustomModel: loss: 2.4398 - accuracy: 0.7283 - f1_m: 1.1505 - precision_m: 142484.9571 - recall_m: 1.1362 - val_loss: 3.0672 - val_accuracy: 0.0990 - val_f1_m: 0.9958 - val_precision_m: 0.9958 - val_recall_m: 0.9958

PreTrained Model: loss: 0.9046 - acc: 0.7174 - f1_m: 0.2450 - precision_m: 301254.4277 - recall_m: 0.2275 - val_loss: 1.0795 - val_acc: 0.6811 - val_f1_m: 0.2366 - val_precision_m: 248592.8750 - val_recall_m: 0.2270

Virtual Reality

For handling virtual reality portion of the project I am going to use WebXR. This doe not require installation of application on the phone. It supports various platforms like oculus and android

Model Architecture

That's all for today. Meet y'all later

So long

Tl;dr: Model dev cycle complete

References

MobileNetV2

WebXR

CustomModel

PreTrained

 Community Bonding Period -

What progress I made this week?

New Learning -

Community Bonding Period June 1 to June 6

Coding period is about to begin tomorrow & I have done preparations to start coding, During this week I had completely set up DHIS2 with program rules/Stages

Few thing I want to Discuss

Baby Classification Logic

Earlier I thought to do Classification in app only , but now I learned about program rules which run automatically and can classify baby easily, Moreover these rules are very much configurable so users can configure these according to their requirements

Currently I used ECEB Action plan to classify Baby’s health in 3 categories - Normal/Problem/Danger

Example of Classification Logic For Danger

#{ECEB_Severe_Jaundice} || #{ECEB_Assess_Temperature}<= 35.5 || #{ECEB_Assess_Temperature} >= 37.5 || #{ECEB_Weight} < 1500 || #{ECEB_Chest_Indrawing} ||#{ECEB_Not_Feeding} || #{ECEB_Fast_Breathing} || #{ECEB_Convulsions}

If any of the above condition is true then the health is in danger.

The program rule has an acion - It classifies health by filling data element ECEB_Classification after evaluating the above expressions.

Creating an Initial Screen to fetch all data elements/program ids from dhis2

For doing API call related to a program/event we need its unique id beforehand, As these ids are not constant on every DHIS2 , I need to first design a screen which fetches all the variables/Stage/Programs ID from the dhis2. Then the user will be able to use the app.

On Call Doctor’s Program

This is an Event Program, Doctor’s schedule is added as an event with data elements name, date, start time, end time, Doctor’s Image.

What progress I made this week?

• Created program rules for Normal/Problem/Danger
• Created On Call Doctor’s Program
• Created Stage 4 & Stage 5 in DHIS2 & modified previous stages according to ECEB action plan

What I have planned to do next week?

• Developing Register baby screen UI
• Developing Assessments Screen UI of App
• Adding Dark Mode in App
• Adding navigation drawer in app with menus ( Doctor’s Schedule, Messaging, Share App, Report Bug)

 Hello everyone, I am glad to share that my project proposal was accepted by LibreHealth in GSoC-21 and I am going to spend my summer working with LibreHealth Community on the Project: Scaling up the mobileHBS/DHIS2 Tracker and Trainer applications — Projects / mHBS — LibreHealth Forums.

I am writing this blog to share my community bonding phase with LibreHealth. It was scheduled from 17th May to 7th June and the coding period will begin in about 10 days.

Project Description

The mobile Helping Babies Survive powered by DHIS2 (mHBS/DHIS2) Tracker (Java) and Trainer (HTML; JavaScript) applications, built within the DHIS2 Mobile platform, are designed to provide educational and training support for maternal and neonatal care among health care providers in low/middle-income countries. We help nurses and physicians in sub-Saharan Africa, Asia, and Latin America deliver evidence-based interventions that save maternal and newborn lives.

In various low/middle-income countries children die at a very low age just in few days after their birth due to the inadequate supply of facilities that they require. The important fact is that all of these deaths are PREVENTABLE and can be prevented by providing proper knowledge to the mother’s of the babies as well as to other person and by providing a proper tracker facility to track the health status of the baby in the early days of his birth.

mHBS (mobile Helping Babies Survive) solve these issues. There are 4 mobile applications, in various stages of design, development, evaluation, and deployment, within the award-winning mHBS/DHIS2 suite of digital tools.

1. mHBS/DHIS2 Tracker (work: data collection)
2. mHBS/DHIS2 Trainer (work: educational and training materials)
3. Essential Care for Every Baby Digital Action Plan (work: clinical decision support)
4. NeoRoo (work: data collection, education, and automated monitoring of neonatal vital signs among babies utilizing the NeoSMART biomedical device)

I will be working on the scale-up of the already existing mHBS/DHIS2 Tracker and Trainer app so that they can be published on the Play Store.

Community Bonding Period

I had used the last days of my community bonding period to set up the project and in getting familiar with its code base and most of the time I did one thing that was getting more idea of the project i.e. Research… Research… Research… and getting familiar with technologies that are going to be used in the coding phase to write a structured code.

We are going to use the dhis2 app codebase for our mHBS Tracker app to collect data from the users and mHBS Trainer app is not a java or android project it is written in HTML, CSS & JavaScript and we are using Cordova for running it on Android Platform.

1. Read articles, documentation related to the project.
2. Learned how to use Cordova for running it’s project on Android.
3. Set up the codebase and ready to code.
4. Learned a lot about the technologies and codebase.

Important Links

1. Talk thread: Proceeding With GSoC
2. Trainer App: GitHub Repo
3. Tracker App: GitHub Repo

Conclusion

It was very great to start with the LibreHealth community.

This week onwards, I will be posting regular weekly blog posts on the updates on further progress on this project here. Thanks for reading! 🙂

Community Bonding Period May 18 to 1 June

Hello Everyone,

I am very happy to start my Google Summer of Code journey with LibreHealth, I will be working on Mobile App Development Project, I am very much Thankful to LibreHealth & Google for selecting me again. I have also successfully completed GSoC previous year & have worked on Cost of Care project of LibreHealth.

About Project

To develop an Android & IOS application to provide clinical decision-support for nurses and doctors delivering essential newborn care interventions during the first day of life. This application provides knowledge, skills, and competencies to nurses and doctors in low/middle-income settings so that they can provide life-saving care to newborns from birth through 24 hours postnatal. App uses DHIS2 as backend & can be setup for any hospital/organization

The Goals of this project are

• Ability to track multiple babies at a time;
• Time-stamping births to generate an essential newborn care clock for each baby being tracked in the app;
• Automated classification of babies health status, based on results from observations and assessments that are conducted during the first 2 hours after birth;
• Automated advice on management of babies, based on their classification; Reminders/alerts to provide recommended essential newborn care interventions;
• If particular interventions are not delivered on time, or at all, capture of information regarding why this essential newborn care was late, or missed

About Community Bonding Period

• Community Bonding Period in which you get to know your community and get familiar with their code base and work style.

• Ask for resources, required by you, to gain your knowledge for the parts of your project you don’t know.

• Discuss and plan your GSoC work with your mentor.

• Discuss and plan your time-table for your summer.

My Progress

• During the community bonding I have introduced myself to mentors & mentors are very much welcoming.
• I recieved access to DHIS2 for the application development, I have discussed with them about creating Programs, Stages, Attributes in DHIS2.
• I read the documentation of DHIS2 Tracker API, Creating Program/Stages.
• I have created a program for On-Call-Doctors & I have to discuss it with mentors.
• I resolved my doubts related to modifying the already created program in DHIS2.

Tasks For Next Week

• Completely setting up DHIS2 with all Programs/Stages/Attributes
• Finalising weekly tasks & creating milestones
• Reading Documentation
• Discussing & showing mentors the programs I will modify in DHIS2

Community-bonding-

Hi, This is Mahavir. This is the very first blog that I am writing during the community bonding period.Today is 31 May and it’s been 13 days since the GSOC result was announced. So let’s have a look at how I spent the community bonding period till now.

My Project-

Bonding-

In the first week of the community bonding period, I introduced myself at the LibreHealth forum. Total 5 students are selected for GSOC 21 program. In this period I communicate with my mentors and other students.

Currently, This project working on the flutter 1.22.6 version, and my first task to upgrade this project to flutter the latest version and null safety. Till now I did some research about flutter's latest version and read about the latest version of dependencies that are used in this project. During this research, I found that GPS dependency is not updated to null safety this was a blockage to upgrade this project.

I discussed this issue with my mentor and muarachmann upgraded this package to null safety.

Conclusion-

overall I spent this time on my learning I learned few new things about flutter's latest version and null safety. I know this is a very short blog and I hope the upcoming weeks are more interesting I am excited for the coding period.

Stay tuned for the next blog.

Happy coding.

Hi My Name is Shivaditya Shivganesh, This is first blog post with regards to Google Summer of Code commonly known as GSoC for the year 2021.

I have been selected for the project Project: Neural network-based object detection of anatomical structures and medical artifacts in Virtual Reality

At present we are in the community bonding phase. Which last for about 15 days the actual coding portion of GSoC starts from 7th of June.

I have started to work on Neural Network model that would be required for the project.

About the Project

In this project we are trying to create a Neural Network model that can. classify the respective classes in the Kvasir dataset and display it in an appropriate VR Display.

The Hyper-Kvasir dataset consists of  images which can be classified into 23 classes.

Conclusion

This blogpost was just for the  community bonding purposes. In the forthcoming weeks I plan to write posts regularly about my progress and the challenges I faced.

Until next time

Hello !

Its been quite a while since the GSoC results were announced and community bonding period commenced. The calendar says May 30 and the coding period will begin in about 10 days. So lets move on to how I actually spent the community bonding period yet and what plans do I have for the coming days before the coding period starts.

Project Introduction

The project bases its goal on creating an EHR system using the webcomponents implementing FHIR resources developed during the previous versions of GSoC. For the implementation some new resource based components will be required as well which will be created simultaneously.

During GSoC 2018, under this project a polymer application was developed which combined all the web component0s (developed at that time) into a PWA.

The current project will be built upon this application by adding and implementing workflows. the workflows that will be implemented in the EHR are :

• Check-in
• Visit
• e-Prescription
• Appointment
• Lab orders

Amazing Mentor Team

1. Saptarshi Purkayastha
2. Namratanehete

Community Bonding Period

Background

During the application period upon the suggestions of the mentors I had decided to mainly work on the CI and the Designs of the PWA during the community bonding period.

Even before the results were announced I was already experimenting with setting up the project so as to identify any problems that may occur as using lit-components inside React can be a little trouble because React never exposes the real DOM.

Bonding Period

So as soon as the results were released I started looking on these parts. I started with the CI part first, one of the pointers in this was also to look for a cross platform - cross browser testing method (SauceLabs) suggested by my mentor. I looked at some of the possibe integrations with these automation tools here.

But most of these are based on e2e testing ( cypress , Puppeteer and others ) and our main focus is on unit and integration testing which will require the react component to be mounted and tested , for which many tools are available but I could not find their integration with sauce or any cross browser-cross platform tool. Once I figure out a way for this issue the CI part will hardly take any time !

the design part:

I have designed one workflow ( check-in ) and will complete the remaining in the coming 2-3 days ! adding to these I will document the endpoints that will be used in each steps in the designed workflows. The endpoints of the check-in workflow is provided below.

here is the video of the designed check-in workflow:

check-in.mp4 from Shashwat on Vimeo.

CHECK-IN WORKFLOW :

SEARCH PARAMS :

GET [base]/Patient?identifier=[value]

GET [base]/Patient?name=[value]

CREATE PATIENT :

POST [base]/Patient

GET PATIENT :

GET [base]/Patient/[_id]

Conclusion

This week’s blog was not too long as I have not written any code, but hoping to deliver some interesting content in the upcoming weeks! I will be posting a blog every week highlighting the things I completed during the week and the things that I plan to achieve.

See you next week 😄

 Defanging an IP Address-

Input: address = "1.1.1.1"
Output: "1[.]1[.]1[.]1"

Input: address = "255.100.50.0"
Output: "255[.]100[.]50[.]0"

 How to append char to end of string in C++ || Best way to append char to end of string-

Method -1 

Method 2-

 Prime Game February long challenge 2021 solution with code and explanation  -

Problem statement -

3
1 2
3 1
2021 42

Chef
Divyam 
Divyam

Hint-

 Team Name codechef  February 2021  long challenge editorial Full solution with code-

Problem Statement-

3
2
suf mas
3
good game guys
4
hell bell best test

2
0
2

 Frog Sort codechef February 2021  long challenge editorial -

3
3
3 1 2
1 4 5
3
3 2 1
1 1 1
4
2 1 4 3
4 1 2 4

3
6
5

Problem Statement- 

 Maximize Function codechef February 2021  long challenge editorial -

Problem Statement- 

3
3
2 7 5
3
3 3 3
5
2 2 2 2 5

10
0
6

Codechef February  long challenge solution 2021-

 Problem Statement-

Input:
N = 3
W = 4
values[] = {1,2,3}
weight[] = {4,5,1}
Output: 3

Input:
N = 3
W = 3
values[] = {1,2,3}
weight[] = {4,5,6}
Output: 0

 Fair Elections Codechef January long challenge  2021 solution || Fair Elections Codechef January 2021 Editorial -

Problem Statement-

Elections are coming soon. This year, two candidates passed to the final stage. One candidate is John Jackson and his opponent is Jack Johnson.

During the elections, everyone can vote for their favourite candidate, but no one can vote for both candidates. Then, packs of votes which went to the same candidate are formed. You know that for John Jackson, there are NN packs containing A1,A2,…,ANA1,A2,…,AN votes, and for Jack Johnson, there are MM packs containing B1,B2,…,BMB1,B2,…,BM votes.

The winner is the candidate that has strictly more votes than the other candidate; if both have the same number of votes, there is no winner. You are a friend of John Jackson and you want to help him win. To do that, you may perform the following operation any number of times (including zero): choose two packs of votes that currently belong to different candidates and swap them, i.e. change the votes in each of these packs so that they would go to the other candidate.

You are very careful, so you want to perform as few swaps as possible. Find the smallest number of operations you need to perform or determine that it is impossible to make John Jackson win.

Input

• The first line of the input contains a single integer TT denoting the number of test cases. The description of TT test cases follows.
• The first line of each test case contains two space-separated integers NN and MM.
• The second line contains NN space-separated integers A1,A2,…,ANA1,A2,…,AN.
• The third line contains MM space-separated integers B1,B2,…,BMB1,B2,…,BM.

Output

For each test case, print a single line containing one integer ― the smallest number of swaps needed to make John Jackson win, or −1−1 if it is impossible.

Constraints

• 1≤T≤1031≤T≤103
• 1≤N,M≤1031≤N,M≤103
• 1≤Ai≤1061≤Ai≤106 for each valid ii
• 1≤Bi≤1061≤Bi≤106 for each valid ii
• the sum of NN over all test cases does not exceed 104104
• the sum of MM over all test cases does not exceed 104104

Subtasks

Subtask #1 (20 points):

• A1=A2=…=ANA1=A2=…=AN
• B1=B2=…=BMB1=B2=…=BM

Subtask #2 (80 points): original constraints

Example Input

2
2 3
2 2
5 5 5
4 3
1 3 2 4
6 7 8

Example Output

2
1

Explanation

Example case 1: We can perform two swaps ― each time, we swap a pack of 22 votes from AA and a pack of 55 votes from BB. After that, John Jackson gets 5+5=105+5=10 votes and Jack Johnson gets 2+2+5=92+2+5=9 votes.

Example case 2: We can swap the pack of 11 vote from AA and the pack of 88 votes from BB. After that, John Jackson gets 8+3+2+4=178+3+2+4=17 votes and Jack Johnson gets 6+7+1=146+7+1=14 votes.

Solution-

This problem can be solved using brute force approach first we sort both array and we will iterate second array from last and first array form first till we get sum1>sum2  .

after swapping if sum1<sum2 print -1

sum1>sum2 print count of swaping 

else print 0

her is my code follow code below.

Join telegram channel for code and editorial.-https://t.me/competitiveProgrammingDiscussion

 Encoded String codechef January long challenge solution || Encoded String codechef January long challenge solution Editorial 2021-

Problem statement-

An encoder encodes the first 1616 lowercase English letters using 44 bits each. The first bit (from the left) of the code is 00 if the letter lies among the first 88 letters, else it is 11, signifying that it lies among the last 88 letters. The second bit of the code is 00 if the letter lies among the first 44 letters of those 88 letters found in the previous step, else it's 11, signifying that it lies among the last 44 letters of those 88 letters. Similarly, the third and the fourth bit each signify the half in which the letter lies.

For example, the letter jj would be encoded as :

• Among (a,b,c,d,e,f,g,h(a,b,c,d,e,f,g,h || i,j,k,l,m,n,o,p)i,j,k,l,m,n,o,p), jj appears in the second half. So the first bit of its encoding is 11.
• Now, among (i,j,k,l(i,j,k,l || m,n,o,p)m,n,o,p), jj appears in the first half. So the second bit of its encoding is 00.
• Now, among (i,j(i,j || k,l)k,l), jj appears in the first half. So the third bit of its encoding is 00.
• Now, among (i(i || j)j), jj appears in the second half. So the fourth and last bit of its encoding is 11.

So jj's encoding is 10011001,

Given a binary encoded string SS, of length at most 105105, decode the string. That is, the first 4 bits are the encoding of the first letter of the secret message, the next 4 bits encode the second letter, and so on. It is guaranteed that the string's length is a multiple of 4.

Input:

• The first line of the input contains an integer TT, denoting the number of test cases.
• The first line of each test case contains an integer NN, the length of the encoded string.
• The second line of each test case contains the encoded string SS.

Output:

For each test case, print the decoded string, in a separate line.

Constraints

• 1≤T≤101≤T≤10
• 4≤N≤1054≤N≤105
• The length of the encoded string is a multiple of 44.
• 0≤Si≤10≤Si≤1

Subtasks

• 100100 points : Original constraints.

Sample Input:

3
4
0000
8
00001111
4
1001

Sample Output:

a
ap
j

Explanation:

• Sample Case 11 :

The first bit is 00, so the letter lies among the first 88 letters, i.e., among a,b,c,d,e,f,g,ha,b,c,d,e,f,g,h. The second bit is 00, so it lies among the first four of these, i.e., among a,b,c,da,b,c,d.

The third bit is 00, so it again lies in the first half, i.e., it's either aa or bb. Finally, the fourth bit is also 00, so we know that the letter is aa.

• Sample Case 22 :

Each four bits correspond to a character. Just like in sample case 11, 00000000 is equivalent to aa. Similarly, 11111111 is equivalent to pp. So, the decoded string is apap.

• Sample Case 33 :

The first bit is 11, so the letter lies among the last 88 letters, i.e., among i,j,k,l,m,n,o,pi,j,k,l,m,n,o,p. The second bit is 00, so it lies among the first four of these, i.e., among i,j,k,li,j,k,l.

The third bit is 00, so it again lies in the first half, i.e., it's either ii or jj. Finally, the fourth bit is 11, so we know that the letter is jj.

Solution-

THis question is based on pattern finding question. You have to find a simple pattern  that 

if(s=="0000"){
	return 'a';
	}
	if(s=="0001"){
	return 'b';
	}
	if(s=="0010"){
	return 'c';
	}
	if(s=="0011"){
	return 'd';
	}
	if(s=="0100"){
	return 'e';
	}
	if(s=="0101"){
	return 'f';
	}
	if(s=="0110"){
	return 'g';
	}
	if(s=="0111"){
	return 'h';
	}
	if(s=="1000"){
	return 'i';
	}
	if(s=="1001"){
	return 'j';
	}
	if(s=="1010"){
	return 'k';
	}
	if(s=="1011"){
	return 'l';
	}
	if(s=="1100"){
	return 'm';
	}
	if(s=="1101"){
	return 'n';
	}
	if(s=="1110"){
	return 'o';
	}
	if(s=="1111"){
	return 'p';
	} Now you have to iterate over string in set of 4 character. finally print ans. follow my code

Join telegram channel for code and editorial.-https://t.me/competitiveProgrammingDiscussion

 Point Of Impact Codechef January 2021 solution || Point Of Impact codechef january 2021 editorial -

Problem statement-