nestauk / ahl_targets Goto Github PK

View Code? Open in Web Editor NEW

0.0 0.0 0.0 160 KB

Exploring the feasibility and health effects of retailer health targets.

Home Page: https://www.nesta.org.uk/project/industry-targets-to-improve-health/

License: MIT License

Shell 0.07% Makefile 1.15% Python 87.11% R 11.67%

ahl_targets's People

Contributors

Watchers

ahl_targets's Issues

explore creating streamlit app

Add dotted line to legend key to figure 6

For different levels of ED and NPM score select examples and give baseline stats

100-200 kcal/100g: an avocado (ED of 150 kcal/100g, NPM is X, not HFSS)
200-300 kcal/100g etc...

Chose examples based on the most purchased or most relatable products
Do 2 for each group

Check what products are excluded when excluding liquids

Write list of most granular category

import utils and major pipeline functions

Simulation to model changes in ED, NPM, sales on kcal/day

create table to summarise scatter plots

function to create histogram of calorie weighted sales for a given brand

function to create histograms of calorie weighted sales for a given food category

summarise results for project kickoff meeting

Determine the impact of healthy high ED foods on targets

function to create bar chart of calorie weighted sale distribution across HFSS, NPM and ED for a given brand

add functions to define whether a product is in scope or not based on HFSS

Repeat figure 5 but subset to HFSS

function to create histograms of calorie weighted sales for a given store category

Update plots of market share of retailers and brands based on new business selection

Repeat charts in slides 22 and 23 https://docs.google.com/presentation/d/1jT4_JYFRExkuovWy0XA2UiC7iAkq7BXHqitVpcx_kc4/edit#slide=id.g244d531d777_1_0

check HFSS shares against retailer published data

Summary statistics (store and brand)

Avg ed per store/brand
Min ed per store/brand
Max ed per store/brand
HFSS per store/brand
% per NPM score per store/brand
Avg sales weighted NPM

Histogram vertical lines repeating

In the script https://github.com/nestauk/ahl_targets/blob/dev/ahl_targets/analysis/histograms.py the vertical lines remain on subsequent plots even when closing the previous one.

Identify the most appealing scenarios according to criteria of: low/medium ED sales increasing, overall calories decreasing, increase in £

Move script to split out hard discounters to pipeline

create options for using different weighting schemes

Quantify impact for retailers (absolute vs relative)

investigate performance issues

Solomon to help with investigation, with a focus on hfss.py
Elena & Izzy to review

Generate one liner for relationship between ED and HFSS, NPM

Using high ED definition as 400kcal/100g and above: what % of high ED products are HFSS? What is the average NPM for high ED products?

change label and title in figure 7

Table of shares for products

Repeat table in the business selection tab of https://docs.google.com/spreadsheets/d/1oEKQAyWFbfIhuATXI6IgZ8l4oY_v6rWelkWaqY16s0I/edit#gid=393481957 but for manufacturers

Add manufacturer info to purchase_complete function in model_data script (pipeline)

Looking at the results it looks like this is info per product and has categories, nutrition, HFSS/npm, some purchase info and store information. Would it make sense to bring in the manufacturer info in this function too? That way its then multi-use for store and manufacturer analysis tables rather than having a seperate function for each.

Update https://github.com/nestauk/ahl_targets/blob/dev/ahl_targets/pipeline/energy_density.py based on https://github.com/nestauk/ahl_targets/blob/dev/ahl_targets/pipeline/hfss.py so that all ml products are converted to g and ED is calculated based on values per 100g

    level_pivot["tot"] = (
        level_pivot["Kilos"] + level_pivot["Litres"] + level_pivot["missing"]
    )

nestauk / ahl_targets Goto Github PK

ahl_targets's People

Contributors

Watchers

ahl_targets's Issues

Recommend Projects

Recommend Topics

Recommend Org