Data drift
drift comparedrift compare

drift compare <json_reference> <json_current>


Description

    Drift of each field between a reference and a current data set, and the missing or new fields

Parameters

    json_reference:   The reference rows - String - required
    json_current:   The current rows - String - required
admin
drift compare "[{\"age\": 25, \"plan\": \"basic\"}, {\"age\": 32, \"plan\": \"pro\"}, {\"age\": 41, \"plan\": \"basic\"}, {\"age\": 29, \"plan\": \"basic\"}, {\"age\": 35, \"plan\": \"pro\"}, {\"age\": 52, \"plan\": \"basic\"}, {\"age\": 47, \"plan\": \"pro\"}, {\"age\": 38, \"plan\": \"basic\"}, {\"age\": 30, \"plan\": \"basic\"}, {\"age\": 44, \"plan\": \"pro\"}]" "[{\"age\": 55, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 61, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 48, \"plan\": \"premium\", \"channel\": \"web\"}, {\"age\": 58, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 66, \"plan\": \"premium\", \"channel\": \"web\"}, {\"age\": 52, \"plan\": \"basic\", \"channel\": \"web\"}, {\"age\": 63, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 57, \"plan\": \"premium\", \"channel\": \"web\"}, {\"age\": 49, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 60, \"plan\": \"premium\", \"channel\": \"web\"}]"
mentdb
{"rows_reference": 10, "rows_current": 10, "drifted_fields": 2, "missing_fields": [], "new_fields": ["channel"], "fields": [{"field": "age", "type": "numeric", "mean_reference": 37.3, "mean_current": 56.9, "ks_statistic": 0.9, "ks_p_value": 1.0825088224475365E-5, "null_rate_reference": 0, "null_rate_current": 0, "psi": 2.104, "drift": "high"}, {"field": "plan", "type": "categorical", "categories": 3, "new_categories": ["premium"], "chi2_p_value": 0.021466325204913872, "null_rate_reference": 0, "null_rate_current": 0, "psi": 1.4192, "drift": "high"}]}

drift compare <json_reference> <json_current> <json_params>


Description

    Drift with parameters: bins (deciles of the reference, 10), psi_moderate (0.1), psi_high (0.25)

Parameters

    json_reference:   The reference rows - String - required
    json_current:   The current rows - String - required
    json_params:   The parameters - String - required
admin
drift compare "[{\"age\": 25, \"plan\": \"basic\"}, {\"age\": 32, \"plan\": \"pro\"}, {\"age\": 41, \"plan\": \"basic\"}, {\"age\": 29, \"plan\": \"basic\"}, {\"age\": 35, \"plan\": \"pro\"}, {\"age\": 52, \"plan\": \"basic\"}, {\"age\": 47, \"plan\": \"pro\"}, {\"age\": 38, \"plan\": \"basic\"}, {\"age\": 30, \"plan\": \"basic\"}, {\"age\": 44, \"plan\": \"pro\"}]" "[{\"age\": 55, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 61, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 48, \"plan\": \"premium\", \"channel\": \"web\"}, {\"age\": 58, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 66, \"plan\": \"premium\", \"channel\": \"web\"}, {\"age\": 52, \"plan\": \"basic\", \"channel\": \"web\"}, {\"age\": 63, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 57, \"plan\": \"premium\", \"channel\": \"web\"}, {\"age\": 49, \"plan\": \"pro\", \"channel\": \"web\"}, {\"age\": 60, \"plan\": \"premium\", \"channel\": \"web\"}]" "{\"bins\": 4}"
mentdb
{"rows_reference": 10, "rows_current": 10, "drifted_fields": 2, "missing_fields": [], "new_fields": ["channel"], "fields": [{"field": "age", "type": "numeric", "mean_reference": 37.3, "mean_current": 56.9, "ks_statistic": 0.9, "ks_p_value": 1.0825088224475365E-5, "null_rate_reference": 0, "null_rate_current": 0, "psi": 2.1979, "drift": "high"}, {"field": "plan", "type": "categorical", "categories": 3, "new_categories": ["premium"], "chi2_p_value": 0.021466325204913872, "null_rate_reference": 0, "null_rate_current": 0, "psi": 1.4192, "drift": "high"}]}
Used in use cases (1 recipe)

    Data scientist › Is the model still valid?
Often used with

    equal