search

tmcd82070 / CAMP_RST

R code for CAMP rotary screw trap platform
1 stars 1 forks source link

Preparation for Release 4.5 #95

Closed jasmyace closed 7 years ago

jasmyace commented 8 years ago

OVERVIEW: A new report that reports passage over forklength groups has been accelerated for development.

RESOLUTION: Set up a list of things to do, so as to maintain timelines and expectations. Release with the new forklength report.

DETAILS: In order to ensure a smooth transition for the next release, the following needs to occur.

  1. Update the R version utilized by the Platform.

    This release occurs in between the previous release and the next involving the new campR package, which requires several new packages. Given the amount of updates between the previous version and the one after this, several changes have occurred in the interim, some of which depend on the new packages. However, in order to prevent having to send a new version of R for all users interested in utilizing the new fork length report, a few minor changes have been walked back in this release. These should not affect estimates of passage. These changes will be reintroduced in the package version tied to the next release.

    :fish: Prior to reducing the scope of the fork-length group report to output only Fall-run data, I checked the sum of the catch included for each run against the output from the other two passage reports. While no fish were lost or gained during the process, i.e., catch was equal for all three, I did observe that for the fork-length report, 9 fish that were previously unassigned became assigned. This was due to these 9 fish having recorded fork-lengths, but no recorded biologically assigned life stages. These are tied to trapVisitID == 159. This behavior appears correct. The zip file contains some information pertinent to this checking. Checking.zip

  2. Complete the Big Looper on all reports. This includes reports with dependencies on updated queries and helper functions. For this round of checking, the American was utilized. All reports that required it used a data range of "2013-01-16" to "2013-06-08".

    a. Estimates production by life stage and run. COMPLETE: Aug 22, 2016

    Note that in this version, this uses the new passageWithLifeStageAssign function, which handles the many options associated with the automatic life stage development.
    b. Estimates production for ALL runs. COMPLETE: Aug 22, 2016 c. View all catch records. (Not checked for this release.) d. Export non-Chinook catch records. (Not checked for this release.) e. Sum Chinook by date. (Not checked for this release.) f. Summarize releases. (Not checked for this release.) g. Plot fork length through season. (Not checked for this release.) h. Plot histograms of fork length (lifeStage = FALSE). (Not checked for this release.) i. Plot histograms of fork length (lifeStage = TRUE). (Not checked for this release.) i. Plot the weekly effort over time. COMPLETE: Aug 23, 2016 j. Estimates production for fork-length "life stage" and Fall run. COMPLETE: Aug 22, 2016

  3. Upload reports to the FTP site, into folder Big Looper Beta Release 4.5 Output in folder For Doug. This is ideally done either early in the morning or on the weekend, when the FTP site generally lacks use.

    I have decided not to send these over this time around. Summary estimates of passage can be gleaned from the Excel worksheet tied to number 4. immediately below. But I can move them over if you want.

    a. American -- KB -- files b. Feather -- KB -- files c. Sacramento (RBDD) -- KB -- files d. Mokelumne -- KB -- files e. Knight's Landing -- KB -- files f. Stanislaus -- KB -- files

  4. Upload all passage results for people to see / analyze. All passage estimates are contained in this Excel spreadsheet.

    allEstsCompare FL 2016.08.22.xlsx

  5. Clean out the R-Interface folder of any old programs, files, etc. no longer needed.

    I haven't bothered to check this. The presence of superfluous files doesn't harm Platform performance. This won't matter after the deployment of the R package campR.

  6. Clean out the Outputs folder of any old testing output.

    I'm only sending over stuff to be placed in the R-Interface folder, i.e., not the whole thing. So this doesn't matter.

  7. Version Control in the Platform. See Issue #55.

    a. Place a zip file containing the release name in folder R-Interface, so the appropriate field ends up populated within the Platform. COMPLETE: Aug 23, 2016

    b. Check that the zip file is working correctly. image

  8. Move all updated programs to the appropriate folder in the local C: drive, for Platform testing. COMPLETE: Aug 23, 2016

    Actually, for this release, all programs originated on my local C: drive.

  9. Test, directly via the Platform, that all reports work as intended. All of the following zip files should have a run_R.out text file, particular to that Platform report run.

    a. Estimates production by life stage and run. LSandRUN.zip b. Estimates production for ALL runs.

    1. day. ALL.Runs.DAY.zip
    2. week. ALL.Runs.WEEK.zip
    3. month. ALL.Runs.MONTH.zip
    4. year. ALL.Runs.YEAR.zip

    c. View all catch records. (Not checked for this release.) d. Export non-Chinook catch records. (Not checked for this release.) e. Sum Chinook by date. (Not checked for this release.) f. Summarize releases.

    1. fall. (Not checked for this release.)
    2. late fall. (Not checked for this release.)
    3. spring. (Not checked for this release.)
    4. winter. (Not checked for this release.)

    g. Plot fork length through season.

    1. fall. (Not checked for this release.)
    2. late fall. (Not checked for this release.)
    3. spring. (Not checked for this release.)
    4. winter. (Not checked for this release.)

    h. Plot histograms of fork length.

    1. lifeStage = TRUE
      1. fall. (Not checked for this release.)
      2. late fall. (Not checked for this release.)
      3. winter. (Not checked for this release.)
      4. spring. (Not checked for this release.)
    2. lifeStage = FALSE
      1. fall. (Not checked for this release.)
      2. late fall. (Not checked for this release.)
      3. winter. (Not checked for this release.)
      4. spring. (Not checked for this release.)

    i. Plot the weekly effort over time. WEEKLYEFFORT.zip j. Mokelumne BYPASS (Not checked for this release.) k. Estimates production for fork-length "life stage" and Fall run.

    1. day. FL.DAY.zip
    2. week. FL.WEEK.zip
    3. month. FL.MONTH.zip
    4. year. FL.YEAR.zip

  10. Sync all current code from the R-Interface folder to the master branch of GitHub.

    This update does not originate from a branch on GitHub.

  11. Move a zipped copy of the updated R code to folder Interim R Releases on the FTP site.

    A zip folder entitled unZipInRInterfaceAndDeleteZIP.zip has been placed in folder Special Release v.4.5 Aug. 23, 2016 for FL in folder For Doug on the ftp site. Its contents should be placed in the R-Interface folder of a Platform folder set-up specifically for this release.

  12. Compile release notes for current release, and formally release the release.

    I have not formally put this as a release; the system wants to tie this to a branch, but since I had to walk things back a bit, a branch tied to this version doesn't exist. Here is the zip file sent to Doug.
    unZipInRInterfaceAndDeleteZIP.zip

At this point, the updated R code is ready for release, and the next stage in the checking process.

jasmyace commented 7 years ago

I don't know why I have several of these release-check Issues in here. I'm closing this one, as it seems unnecessary.