Generating Prospect Likelihood scores
To generate Prospect Likelihood scores, perform these calls:
compute_features.py -i history01.sqlite -o features01.csv -s 2016-01-01 -e 2016-12-31
compute_scores.py -i features01.csv -o scores01.csv -m config.txt -t 0.86 0.935
This table shows the command line parameters of the call:
| Option | Abbreviation | Required | Comment |
|---|---|---|---|
| input | -i | Yes | Input in SQLite format. |
| output | -o | Yes | Output CSV file. |
| startdate | -s | Yes | Configuration file with the transformation format.
See Format of imports. |
| enddate | -e | Yes | Number of lines processed at once. |
| model | -m | No.
Default "," |
Delimiter used for CSV input filet. |
| threshold | -u | No.
Default false |
Data is updated only if changed. False can be used to enforce import if no history data exists. |
| escape | -e | No.
Default "{" |
Escape character inside strings. |
| new | -n | No.
Default false |
Create new SQLite file on import. |