learnings

REPO: basic interaction with db: run scripts, retrieve query results: SCRIPT, QUERY, CONNECTION

Model: a WHOLE table
breaks into......
ModelInstance: individual records: UPDATE, SAVE
DirScan: scan a DIRECTORY: COLLECTION, SAMPLE, how to process FILES
      SasScanner: for sas database files

Find one associated record by type and name


row count, column count, date created

variable: label, type, length

Periodicity: data updated....two months go by.....SAME data updated....two months go by.....SAME data updated

Seeding a database with data, seed.yml file etc.

DETERMINE COLLECTION
agency name: hhs
collection name: tanf
sample name: 2008


a fulltext index is faster than a like or full query cus it looks for a partial piece type thing


SQLServerDataSource  MICROSOFT SQL server


SELECT TABLE_NAME
FROM INFORMATION_SCHEMA.TABLES
WHERE TABLE_TYPE = 'BASE TABLE' AND TABLE_CATALOG='cmr';



Model: what data structure looks like in db, but make a "picture" of it

directory names~agencies

create_repository is just a SQL schema

geography_level is important (some are state samples, potential exists)

SIPP record type is the person_month record, which is the series of all of the months of data for a specific person

records are like our rec_types
record layout is the thing the variables are attached to
if we have 10 yrs hud pik layout (record layout), want to save time, but if they are different then have to use new

samples vs. records vs record layouts

HUD, veterans and wives, 12 yrs, and 5yrs have this skeleton structure (cus they might want to know comparatively what kind of record layout)

minimize repeating


there is seeds agencies and dataCollections HHS has TANF and CCDF

Application: gets command options(a combination of what came in on command line and properties, use properties value as default, otherwise command line, 

using COMMONS CLI library for options stuff

A record has a sample_id
A sample has a datacollection_id and name
on Model there are three ways to get instance- initialize, findby name, findbyId
SasFileProperties

"Was it a new one, was it an UPDATE"

get rest of db working and thing parsing (SAS guy) 

HELP
-making seed complete with datasets
-see if i can reproduce it we can work on different datasets

Comments

Post a Comment

Popular posts from this blog

some url/category/action_within_category create an instance of the category controller and then calls the action with that instance looks for template to display the action and result heading to /views/say and seeing if it finds a specific template for that action
Read more
a bunch of methods that can be used with any other class (MIX IN) a method is a group of several expressions(block) that can be referred by a name, passed arguments, associated with one or more objects higher order function takes a function as a param or returns as a result AREL relational algebra  returning the reply for the first comment, a table joining itself why is that easier in arel than sql, because sql qould require aliasing CONTEXT When    user logs out when    user logs in when    not valid or With TEST VALID EDGE NOT VALID ex  describe destroy do        context 1: WHEN resrouce found => 200        context 2: WHEN resource not found        context3 : WHEN resource not OWNED expect DONT SHOULD expect dont should expect dont should expect means in my mind i think i know what will come rather than should, if what comes isn't what i thought obl...
Read more

ADVERTISING STYLES

storytelling      lifestyle focus MARKETING OBJECTIVES AWARENESS Brand - those MORE likely to pay attention to ads and increase awareness Local - geographic proximity Reach - Max # ppl CONSIDERATION traffic - send more people to a destination ON or OFF facebook engagement - comments, shares, likes, event responses, offer claims app installs -  video views lead generation - collect lead info from ppl interested in your business CONVERSION conversions - valuable actions (ex. adding payment info, making purchase) product catalog sales - automatically show products from catalog based on target audience store visits - promote geographic business location to those nearby AUDIENCE SELECTION Geographic location age gender language Detailed targeting include/exclude :  DEMOGRAPHICS INTERESTS BEHAVIORS Connections Facebook pages - people who like, friends of people who like, exclude Apps - same as above Events - same as...
Read more