Imputation

Make sure you've generated the graph db. See Graph Generation

Make sure the virtual environment is active.
```
 cd grimm
 source venv/bin/activate
```
Prepare data The don.gl.txt.gz file is generated by g2gl method. See WMDA g2gl
```
 cd validation/wmda/data
 gunzip -c don.gl.txt.gz > don.gl.txt
 gunzip -c pat.gl.txt.gz > pat.gl.txt
```
If you need to reproduce consensus results for WMDA MVT 3 exactly, the donor D004016 needs to be updated with 33:03 typings. Apply the change:
```
 git apply don.gl.txt_mvt-3-discrepancy.diff
```

Install gl-impute module.

 cd ../../..
 cd gl_impute
 make install

Update neo4j.json file with the username/password for your Neo4J instance if different.
```
 cd ../multi_race_impute
 edit neo4j.json
```
Start the imputation generation process. This process currently takes 2+ hours.
```
 time python RunFile.py
```
.... wait ....
```
 	175.96s user 6.40s system 2% cpu 2:25:50.00 total
```

Finally.. The imputation results are produced in output directory.

 output
 ├── don.gl.txt_miss
 ├── don.gl.txt_out
 ├── don.gl.txt_val
 ├── pat.gl.txt_miss
 ├── pat.gl.txt_out
 └── pat.gl.txt_val

Provide feedback