Currently the edge durations are based on any subsetted dates, but instead should use the full range of dates available for the roster data even if degree/mixing/other cross sectional network stats are within a more limited date range

This is in step 15

Default does not require any external libraries to be installed...

e.g., the FCJ Network Analysis.Rmd

Can the two RDS files go in a data/output folder?

If there are other considerations for classifying how residents interact, please let us know. For example, Anne, you mentioned the trusted residents who deliver food, etc. If there is some way to identify them on the rosters, that could help.

****Trustees are housed in a particular cell block. For example, they may all be in 1 North 100 zone. I need to check if they are *always * in this zone, or if the designation of which zone is the trustee zone for meal deliverers changes. Can I get back to you about this?

• Overall rate of intake into the jail
• Rate of movement between floors/zones/cells within the jail
• Movement of rates stratified by floor (as lower floors are higher turnover)
• Degree distributions (counts of 0, 1, 2, 3....) by network layer type, at different time points
• Above stats stratified by demographics (age, sex, race)

Avoid the many warning messages while rendering the main Rmd:

`summarise()` has grouped output by 'head.age'. You can override using the
`.groups` argument.

Ensure package versions are consistent across users

1. There are four SO Numbers in the dataset that don’t start with ‘P’ and leading zeros (543146, 618354, 711911, 712979). Is there something different about these four residents?

• 543146 (female) appears once in the data, on 01/12/2021
• 618354 (female) appears several times in the data, from 01/11/2022 to 02/04/2022
• 711911 (male) appears once in the data, on 10/27/2021
• 712979 (male) appears several times in the data, from 10/27/2021 to 02/04/2022

2. 2 cells are marked as ‘Padded’ (3LD60 on 10/27/2021 and 3LD82 on 01/16/2022). Are these cells geographically separated from the other 3LD[#][#] cells? Are there other padded cells in the data that are not marked as such?
3. Robertson, Freddie Jerome (SO Number P01007978) has Gender = “M” from 10/27/2021 to 01/31/2022 and Gender = “F” on 02/04/2022. Is this intentional?
4. Certain locations (listed below) don’t fit into the Floor + Tower + Block + Cell pattern. How do they fit into the architecture of the prison? Where are they located relative to the North/South towers?

• Central Holding, Central Release, Intake, Medical Holding, Weekender (Intermittent), In Transit Cell. These locations appear throughout the data on various dates, for both male and female residents
• 3LD[#][#] (e.g., 3LD30). These locations appear throughout the data on various dates, for both male and female residents
• [Letter][#][#] (e.g., A20). These locations appear throughout the more recent data (from 01/12/22 onwards) for female residents
• H[#] (e.g., H3). These locations appear throughout the more recent data (from 01/12/22 onwards) for female residents
• MA[#][#][#] (e.g., MA217). These locations appear throughout the more recent data (from 01/12/22 onwards) for male residents

Can embed the analysis with:

`{r xxxxjjjjf}`

Tables

Output data as CSV, then copy'n'paste into clean Excel file (template on OneDrive).

• T1: Basic descriptive stats of jail and residents. You have many of these already in your Rmd. We can tweak what should go in here later.
• T2: Summary stats of mean degree, age mixing with columns for overall, January, and April months. Averaging over the entire 11 day string for each month.
• T3: Dissolution/duration stats

Figures

For R outputs, please generate these as high-resolution TIFF files and also PDF files, in OneDrive

• F1: network schematic
• F2: PRISMA diagram
• F3: Box plot (two panels, daily degree for January string of dates and April string of dates. Remove any dates before or after or in-between. Use a clear color for the boxes to differentiate the month. Two panels, cell and block networks. Example: Figure 2 here
• F4: Heat map/contour plot of age mixing, four panels: proportional (standardized) age mixing matrix. Use interpolation so the contour is smoothed. Two on top row: January cell and block mixing; Two on bottom row: april cell and block mixing. Try to use standard legend (color values). For these, average over the entire 11 day string for each month.

> remove(i, j, races, floors, ages, g, bNWs)
Warning in remove(i, j, races, floors, ages, g, bNWs) :
  object 'bNWs' not found

doesn't appear to be a problem, but code needs update