asreview / asreview-datatools

Tool to preprocess datasets for ASReview
MIT License
19 stars 13 forks source link

Add Time to Discovery to output #6

Closed J535D165 closed 3 years ago

J535D165 commented 3 years ago
jonathan$ python -m asreview stat tests/output/ptsd.h5
************  ptsd.h5  *******************

-----------  general  -----------
Number of runs            : 1
Number of papers          : 6189
Number of included papers : 43
Number of excluded papers : 6146
Number of unlabeled papers: 0
Number of queries         : 6188

-----------  settings  -----------
model             : nb
query_strategy    : max
balance_strategy  : double
feature_extraction: tfidf
n_instances       : 1
n_prior_included  : 1
n_prior_excluded  : 1
mode              : simulate
model_param       : {'alpha': 3.822}
query_param       : {}
feature_param     : {'ngram_max': 1, 'stop_words': 'english', 'split_ta': 0, 'use_keywords': 0}
balance_param     : {'a': 2.155, 'alpha': 0.94, 'b': 0.789, 'beta': 1.0}
abstract_only     : False

-----------    ATD    -----------
 0.0165

Time to discovery:

    row   : value
    3898  : 22.0
    675   : 23.0
    284   : 24.0
    592   : 25.0
    1446  : 29.0
    335   : 31.0
    5054  : 40.0
    5244  : 46.0
    4011  : 47.0
    2408  : 50.0
    719   : 51.0
    720   : 52.0
    1568  : 63.0
    2472  : 70.0
    1425  : 73.0
    896   : 79.0
    897   : 81.0
    3489  : 84.0
    2455  : 91.0
    4250  : 93.0
    3053  : 95.0
    4435  : 97.0
    4434  : 98.0
    2616  : 100.0
    4334  : 102.0
    3141  : 103.0
    4938  : 108.0
    4939  : 110.0
    5284  : 115.0
    1933  : 118.0
    4313  : 122.0
    5655  : 130.0
    2446  : 132.0
    2445  : 133.0
    2444  : 134.0
    2382  : 140.0
    4768  : 148.0
    1922  : 151.0
    4104  : 157.0
    3483  : 169.0
    5479  : 234.0
    3316  : 523.0

-----------  WSS/RRF  -----------
WSS@95 : 92.51 %
WSS@100: 91.55 %
RRF@5  : 97.62 %
RRF@10 : 100.00 %
Rensvandeschoot commented 3 years ago

super nice! Will the results be available in a separate json (or csv) file? And would it be possible to add the title as a column next to the record_id?

J535D165 commented 3 years ago

Will the results be available in a separate json (or csv) file?

Stored in the same JSON-file.