1. 25 Apr, 2020 2 commits
  2. 20 Apr, 2020 1 commit
    • Chris Jewell's avatar
      Changes: · ed4c856a
      Chris Jewell authored
      1. Added new data sample file.  New data is incremental so tidied out
         old data to save space.
      2. Added configuration parameters for posterior, prediction, and rt files.
      ed4c856a
  3. 17 Apr, 2020 3 commits
  4. 09 Apr, 2020 2 commits
  5. 06 Apr, 2020 1 commit
    • Chris Jewell's avatar
      Changes: · 9750447b
      Chris Jewell authored
      1. Working MCMC accepting spatial data.
      2. New spatial data 149 UTLAs in England
      2. New case data format added:
          * case data as csv with headers <specimen_date>, <UTLA_code>, <Age>
          * specimen_date is in d/m/Y format (blame upstream providers)
          * age is 5-year age group (age//5 * 5)
      3. Prediction code now has Rt calculation.  However this seems highly inefficient!
      9750447b
  6. 05 Apr, 2020 5 commits
  7. 03 Apr, 2020 1 commit
  8. 02 Apr, 2020 1 commit
  9. 01 Apr, 2020 3 commits
  10. 31 Mar, 2020 1 commit
  11. 30 Mar, 2020 5 commits
  12. 28 Mar, 2020 5 commits
  13. 27 Mar, 2020 6 commits
    • Chris Jewell's avatar
      Merge pull request #3 from csuter/stochastic · 5995a91c
      Chris Jewell authored
      Add requirements.txt file
      5995a91c
    • Christopher Suter's avatar
      Add a simple requirements.txt file. · 56562073
      Christopher Suter authored
      Excludes TF and TFP since we may want to switch between stable and nightly versions
      56562073
    • Chris Jewell's avatar
      Merge pull request #2 from csuter/stochastic · 3cf012e6
      Chris Jewell authored
      Rewrite Multinomial as hand rolled iterated binomial.
      3cf012e6
    • Christopher Suter's avatar
    • Chris Jewell's avatar
      Merge pull request #1 from csuter/stochastic · 79f51e8d
      Chris Jewell authored
      Improve efficiency of stochastic model.
      79f51e8d
    • Christopher Suter's avatar
      Improve efficiency of stochastic model. · 864fcf65
      Christopher Suter authored
      Big changes:
       1. replace python for loop with tf.while_loop
       2. work with a transposed state tensor shape
         - instead of [4, nlads * nages], use [nlads * nages, 4]
         - this made it pretty easy to eliminate some transposes in
           propagate_fn (there were comments there seemingly contemplating
           this shape arrangement)
         - this feels a little more natural to me, too; in TFP we'd call the 4
           SEIR states components of the "event shape" of the system, and the
           nlads * nages part a "batch shape" (although one could reasonably
           also combine these together into one big matrix "event shape")
         - anyway, this allowed elimination of 3 transpose ops which makes for
           simpler code and avoids some memcpys
         - I also made an effort to update surrounding code to use the same data
           layout, but it seems like mcmc.py and covid_ode.py are broken right
           now anyway, due to other changes made in support of stochastic mode,
           so I couldn't confirm that my changes were sufficient.
       3. switch off XLA (which didn't yield any clear improvement, although
          it also didn't really hurt), and disable autograph (which tries to
          do things like rewrite python for loops into TF graph code but tends
          to produce less performant than manually optimized code like what
          I've done here)
      864fcf65
  14. 26 Mar, 2020 4 commits