usepa / useeio Goto Github PK

Impact factors are currently off in https://github.com/USEPA/USEEIO/blob/master/useeiopy/Model%20Builds/USEEIOv1.1/USEEIOv1.1_LCIA.csv

I'm actually leaning toward just moving the loc and root to after the version number and just making USEEIO implicit so we only use it in contexts when its not clear. That is because all these models are USEEIO family models so my earlier change to make these two components part of this core name is problematic

Originally posted by @WesIngwersen in #11 (comment)

occurs with 2012 data.

https://github.com/USEPA/USEEIO/blob/c4eb1700cdaf1345a584a81dbf94e74f2e3f8681/import_factors_exio/useeio_imports_script.py#L494C49-L494C49

This script is designed to generate import factors

1. over a span of year(s)
2. at a specified BEA level of detail

for 1, these parameter should either be clearly the the top of the file with the other params like the paths and not in this statement, or they should be required as params to be passed to this script upon execution

for 2, it's not that clear yet that this script only works at the Summary level which it appears to do. This needs to be very clear in the notes that it only works for that level.

https://github.com/USEPA/USEEIO/blob/12ff0aaa0cf8d56429c95372dc60c568d014ad8a/import_factors_exio/API/Census_API.yml#L3C13-L3C61

I tried to install this from PyPi using pip

pip install currency_converter

It installed the package currency.converter

Upon importing it for the first time i got an error because it depended on zope

I installed zope via pip but it is a whole application server and had a ton of dependencies.

I modified the import line to read

from currency.converter import CurrencyConverter
but the latter is not an object of the former so that didn't work.

Rather than fix this, can we find another package that doesn't have so many dependencies and has been update

FYI @bl-young

The core models we use in USEEIO use the "before redefinitions" tables. TiVA data are only available "after redefinitions". Will this impact the values for some commodities?

https://github.com/USEPA/USEEIO/blob/c4eb1700cdaf1345a584a81dbf94e74f2e3f8681/import_factors_exio/useeio_imports_script.py#L88C64-L89C52

for 2012 it appears 83 of 9800 rows which are by EXIO commodity and EXIO sector get dropped, which is not a high percentage, but still I I don't think this is justifiable.

https://github.com/USEPA/USEEIO/blob/c4eb1700cdaf1345a584a81dbf94e74f2e3f8681/import_factors_exio/API_Imports_Data_Script.py#L236C11-L236C11

FileNotFoundError: API key required for BEA_API. Create the file "../import_factors_exio/API/BEA_API_key.yaml" and add your API key

what is an example format for this yaml file containing the key?
This requirement for a key and the format of this file is not documented

For a MAC R user, ModelBuildspath variable in .Rprofile was not being loaded as it should have.

Right now we get this characterized matrix ..this is equivalent to our N.

But there is also satellite.M
There are many flows for the major GHGs in satellite.M which is perhaps why we choose a characterized version, however, we are missing some important flows like. We should probably get the unit data as well.
We can use grep on all the flows and get those starting with the major GHGs.

If we use satellite.M we can also get SF6, HFCs (as a group) and PFCs (as a group). It's more complete than the three gases that we are getting now.

Under the OpenLCA schema, all objects should be categorized. Flows that are elementary have elementary categories; however, many flows are not categorized. See the following two examples, chosen at random from a recent build:

{
  "@type": "Flow",
  "@id": "0a69c698-a3b0-3945-884c-0a75e5f4a444",
  "name": "1,4-Dichloro-2-butene",
  "flowType": "ELEMENTARY_FLOW",
  "flowProperties": [
    {
      "referenceFlowProperty": true,
      "conversionFactor": 1.0,
      "flowProperty": {
        "@id": "93a60a56-a3c8-11da-a746-0800200b9a66"
      }
    }
  ],
  "category": {
    "@id": "42c9fee0-d374-30e8-b2c5-32fa901a691e"
  }
}

{
  "@type": "Flow",
  "@id": "946a7ad3-1042-3ded-8a7f-8368643372ac",
  "name": "Compensation of employees",
  "flowType": "ELEMENTARY_FLOW",
  "flowProperties": [
    {
      "referenceFlowProperty": true,
      "conversionFactor": 1.0,
      "flowProperty": {
        "@id": "3bf53920-157c-4c2f-bddd-7c92c9d35f10"
      }
    }
  ]
}

https://github.com/USEPA/USEEIO/blob/c4eb1700cdaf1345a584a81dbf94e74f2e3f8681/import_factors_exio/useeio_imports_script.py#L332C4-L332C43

Labeling this as bilateral trade is not clear - which direction, with whom?
labelling as exports to US would be much more clear

JSON-LD spec requires objects to include @context with something resembling the following contents:
"@context":"http://greendelta.github.io/olca-schema/context.jsonld"

https://json-ld.org/
https://github.com/GreenDelta/olca-schema/blob/1ab2f1a52b8a85ab2c81e9fcaaebcbb4db38df20/examples/unit.json

Reported by Zia Chowdury.
EUTR factor for air/unspecified/nitrogen oxides is 0.08.
The correct value from TRACI 2.1 is 0.04429.

These three minor countries are the only contributors to these BEA regions. I'm a bit confused as to why that is and what that does to our RoW emission factor. Note that AsiaAndPac shows up in a contributions by region here, but I doubt that Wallis and Futuna is a good proxy for that

https://github.com/USEPA/USEEIO/blob/c4eb1700cdaf1345a584a81dbf94e74f2e3f8681/import_factors_exio/useeio_imports_script.py#L324C6-L324C6

Please update imports and IO data to work alternatively with newer schema data and update the data.

There are significant imports of Other goods and some of Used goods (see rows from t_df from 2012 these are in million current USD)

but they get dropped after this merge because they are not in the concordance.

These would account for substantial embodied emissions and need to be handled.

I don't think this is needed here

          > @Jnamovich where did the census NAICS -> BEA mapping come from? it's missing some NAICS that only exist in 2012 but no longer exist in 2017

Good catch. Looks like those data were using 2012 NAICS until 2019 when they switched to 2017 NAICS. If that switch is not described in the data its something we should report to the API maintainers and maintain a config file mapping the source by data year to the right NAICS year and make sure our crosswalk has both like we have here in flowsa under NAICS timeseries

Originally posted by @WesIngwersen in #25 (comment)

INFO Renaming sectors to USEEIO names in satellite tables...
Traceback (most recent call last):
File "useeiopyExample.py", line 9, in
useeio1pt1 = useeiopy.assemble("USEEIOv1.1")
File "c:\users\wingwers\useeio-public-test-jun7\useeiopy_init_.py", line 63, in assemble
Satellite.renameSatelliteSectors(satcsvfiles, sattablepath)
File "c:\users\wingwers\useeio-public-test-jun7\useeiopy\Satellite.py", line 61, in renameSatelliteSectors
names = read_names(Crosswalkrelpath + '/MasterCrosswalk.csv')
File "c:\users\wingwers\useeio-public-test-jun7\useeiopy\Satellite.py", line 78, in read_names
names[row[6].strip()] = row[8].strip()
IndexError: list index out of range