repronim / reproschema-ui Goto Github PK

Lets make a list here of the UI components we want to support and their label names.

• radio buttons inputType: 'radio'
• text input inputType: 'text'
• audio record inputType: 'audioRecord'
• integer input inputType: 'xsd:int'
• number input inputType: 'number' (this can be ints or floats)
• date picker inputType: 'date'
• checkboxes ??
• place input
• multitext input
• audioRecordNumberTask

@sanuann can you continue this list?

add instructions:

stop once you are done speaking. You may hit play to hear what was recorded.```

to make it easier for schema writers to know what inputtype to use, there should be a list provided by schema-ui and this should be consistent with the validator in reproschema.

after filling year of birth, if you move to a different activity and return the value is not updated to reflect the entered value.

i don't know how easy it would be to do both pause and play from here on out, or pause and play from the beginning. once recorded, could the soundcloud ui be used (if it's available as a widget)?

and start from https://github.com/sanuann/schema-standardization/blob/master/volumes/ReproNim/nda-phq9.jsonld

after signing the consent it does not proceed.

As I am working on the readme for this, by reinjecting some of my own doc notes and some of the content of other issues (#133 #4 #18) I realize that this does not have a license.

Should I add one in the [DOC] PR I am working on? Apache 2.0 like for the reproschema?

the values for startedat and endedat are not timestamps.

this should allow any non-schema pieces to reflect language switches.

an example vue library: https://kazupon.github.io/vue-i18n/ would be good to do some search on this.

1. select dropdown needs to be fixed in demographics.
2. fix the error in console (not sure why it occurs)
3. other questions assessment - branching between the items
4. branching in voice task feedback questions.

#47

https://app.asana.com/0/774967722873529/1118923825279215/f

reference: https://www.youtube.com/watch?v=Tm3fD89dqRE

When this happens:

Then this appears

But if you click on "show result" nothing happens, it just goes to the next activity.

Option A) we could run a signal-to-noise ratio test (SNR) on that audio sample to see if the environment is too noisy. Then we can give feedback: "Your environment is too noisy, please go somewhere quieter and try again".

So we'd need to run some code to run the SNR and depending on the score either proceed or give feedback, is that possible?

Option B) A simpler option is what you have now, and it's just playing the audio back to them and instructing "If you think there is too much background noise, please go somewhere quieter".

@satra , what do you think?

console logging should be a debuggable option as either a url parameter or a config variable.

• play button should turn to stop when playing and back to play when playing is done
• on android, playing appears to stop shorter than it should
• when you click redo recording, the time limit changes to 10000s from 10s

its not showing the passage!

load from https://github.com/akeshavan/schema-standardization/blob/master/activities/PHQ-9/phq9_schema.jsonld#L67

• keep previous answers highlighted when new item is loaded
• hitting skip at any time, highlights the skip for that item and unselects any selected answer
• selecting an answer on a skipped question, removes skip highlight

for both of the last two cases the answer dictionary is updated to reflect current state.

• something that indicates progress perhaps a bar on top of the screen that stays as you scroll up and down.

https://app.asana.com/0/774967722873529/1118923825279214/f

So multiple choices can be selected

e.g. remove skip/don't know

• should be able to override skipping at different levels

Read the new schema from ReproNim/reproschema#55 and render the titles in the sidebar

1. in the results, two records are being created for each item (including two wav files for each audio item).
2. the attributed to field should generate a unique uuid for a participant (common across all items in a session).

currently the reproschema-ui (schema.repronim.org/ui/) does not load the demo voice protocol from schema.repronim.org/rl

also it should have a demo banner on any screen.

@sanuann @satra

I was going through the checklist for the PET guidelines and the time range widget seems to have some hard coded aspects that won't be usable for the needs for that use case.

I suspect though this might require some changes on the schema side to allow more flexibility, no?

<template>
  <div class="TimeRangeInput container ml-3 pl-3">
    <b-row class="mt-2 mb-2">
      <b-col class="mt-2 mb-2">
        went to bed:
        {{yesterday}}
        <vue-timepicker v-model="slept" format="hh:mm A"></vue-timepicker>
      </b-col>
      <b-col class="mt-2 mb-2">
        woke up:
        {{today}}
        <vue-timepicker v-model="woke" format="hh:mm A"></vue-timepicker>
      </b-col>
    </b-row>

    <div class="mt-2 mb-2" v-if="timeSlept != null">
      <div class="mt-2 mb-2" v-if="timeSlept < 0">
        <b-alert show variant="danger">You can't sleep negative hours!</b-alert>
      </div>
      <b-button v-else @click="sendData">Submit {{timeSlept}} hours</b-button>
    </div>
  </div>
</template>

@sanuann @satra

This might be cryptic way to describe this but for MRI and PET we are getting a lot items that require answer in terms of

X= ; Y= ; Z=

Very often for voxel dimension, image resolution...

It does not seem like a good idea to have 3 separate "float" items for this.

Did I miss it or is this something covered by the schema at the moment on the UI part in terms of possible inputType?

should behave as follows:

1. initialize the widget with default UUID.
2. option to replace with personal UUID.
3. when url is called with particpant_id, the widget is initialized with the participant_id

visibility of widget:
if url is called with participant_id => do not show the widget
else => read only.

allow list to allow this widget to be readonly.

@sanuann

zip each blob in responses so that less data is sent

http://schema-ui.anisha.pizza/?url=https://raw.githubusercontent.com/ReproNim/schema-standardization/master/activity-sets/example/nda-phq.jsonld

Hello,

quick question from total noob: is it possible to serve this schema-ui repo using github pages?

trying to build a standalone proof of concept website so I figured this would the first easy solution.

Thanks

Remi

1. add a new folder here: https://github.com/ReproNim/schema-ui/tree/master/src/components/Inputs (called "date")

2. https://github.com/ReproNim/schema-ui/blob/master/src/components/InputSelector/InputSelector.vue#L43

maybe try this UI:
https://www.npmjs.com/package/vuejs-datepicker

i think the latest updates to results should resolve this issue.

https://app.asana.com/0/774967722873529/1118923825279217/f

We are making a dangerous assumption here: https://github.com/ReproNim/schema-ui/blob/master/src/components/Survey/Survey.vue#L217

That the context prefix we are using maps directly to the name of the jsonld file. We need this because the prefix is how we are defining the scoring and branching logic. @satra when we do jsonld.expand on an activity, we lose the prefixes that we defined in @context -- do you know of a way to recover them?

all parts of name (first, middle, last) in NDA items to be in a single widget

when a user re-records on safari, the audio record does not collect new data.

on mobile devices, it is rather unintuitive right now how to get started with responding. the flow should immediately lead people to go through each question and respond, including moving automatically from one section to the next.