add firework creation time to final json/prediction step

[zulissi]

This would allow us to sort by the discovered materials (i.e. find the ones that are currently new or rising to the top)

[ktran9891]

I thought we had that in the adsorption collection. Example: this Mongo document has a creation time of "ctime": "2017-08-25 12:19:23.067000" and a modification time of "mtime": "2017-08-25 12:19:23.067000".

[zulissi]

I think that's the time it was added to the adsorption collection (which happens periodically when we rebuild the DB), rather than when the calculation ran.

[ktran9891]

Good point. The FireWorks "created on" and "updated on" items (below) are more appropriate.

So we want to put those values into the MongoDB?

ktran@cori06:~$ lpad get_fws -i 27953 -d more
{
    "name": {
        "mpid": "mp-1018179", 
        "calculation_type": "slab+adsorbate optimization", 
        "shift": 0.07726067999999997, 
        "top": false, 
        "user": "zulissi", 
        "adsorption_site": "[  3.57  -4.6  -10.96]", 
        "slabrepeat": "(1, 1)", 
        "miller": [
            1, 
            1, 
            2
        ], 
        "num_slab_atoms": 0, 
        "vasp_settings": {
            "pp": "PBE", 
            "isym": 0, 
            "kpts": [
                4, 
                4, 
                1
            ], 
            "encut": 350, 
            "pp_version": "5.4", 
            "isif": 0, 
            "ibrion": 2, 
            "gga": "RP", 
            "ediffg": -0.03, 
            "nsw": 200, 
            "lreal": "Auto"
        }, 
        "adsorbate": "CO"
    }, 
    "launches": [
        {
            "fworker": {
                "category": "", 
                "query": "{\"name.vasp_settings.gga\": \"RP\"}", 
                "name": "arjuna GPU", 
                "env": {}
            }, 
            "trackers": [], 
            "ip": "10.4.11.27", 
            "fw_id": 27953, 
            "state": "COMPLETED", 
            "host": "c026", 
            "launch_dir": "/home/zulissi/fireworks/block_2017-05-15-22-09-44-974650/launcher_2017-05-16-08-39-11-602428", 
            "action": {
                "defuse_workflow": false, 
                "update_spec": {}, 
                "mod_spec": [], 
                "stored_data": {
                    "opt_results": [
                        "Atoms(symbols='Al4Au8H', pbc=True, cell=[[4.76571695, 0.0, 0.0], [0.0, -5.12846057, 0.01989435], [0.0, 0.0, -29.54066953]], constraint=FixAtoms(indices=[0, 1, 4, 5, 6, 7, 9]), calculator=SinglePointCalculator(...))", 
                        <removed for brevity>, 
                        -39.42954158
                    ]
                }, 
                "exit": false, 
                "detours": [], 
                "additions": [], 
                "defuse_children": false
            }, 
            "launch_id": 22628, 
            "state_history": [
                {
                    "updated_on": "2017-05-16T09:35:22.682729", 
                    "state": "RUNNING", 
                    "created_on": "2017-05-16T09:27:38.255408"
                }, 
                {
                    "state": "COMPLETED", 
                    "created_on": "2017-05-16T09:35:22.829653"
                }
            ]
        }
    ], 
    "fw_id": 27953, 
    "state": "COMPLETED", 
    "created_on": "2017-05-16T08:56:53.388554", 
    "updated_on": "2017-05-16T09:35:24.034953"
}

[ktran9891]

More notes from a meeting: One of the benefits of this would be that we could start tracking the number of hits over time. This can help us gauge the performance of GASpy as a screening tool.

[ktran9891]

This commit fixes this.