tabMWRacc() issues

[ben-wetherill]

Testing tabMWRacc(type = "individual"):

• For Field Blanks and Lab Blanks, if the Quantitation Limit column has a value and the result value is BDL, then the Quantitation Limit should override the MDL from the DQO file. If the result value is not BDL, then the Quantitation Limit should be ignored.
• For Lab Duplicates, I'm getting an error when running the function on TP values that are lower than 0.05 mg/L. If I create TP lab duplicate values of 0.04 and 0.01, I get the following error: Error: Problem withmutate()inputHit/Miss. x <text>:1:5: unexpected numeric constant 1: 120 0.01 ^ i InputHit/Missisdplyr::case_when(...). i The error occurred in row 20.
• For Field Duplicates I'm getting the same error when I create a TP field duplicate of 0.03 and 0.06.
• I also run into a similar error when running for type="percent" and I add a pH lab duplicate for values of 7.2 and 7.25.
• For Field Duplicates, the RPD doesn't seem to be calculated correctly when the result value is AQL. Example from results file: E.coli, RVM-005, 9/13/21, result AQL, QC Ref 450, Quantitation Limit is 1200. RPD should be 15%.
• For Field and Lab Blanks, the threshold is not always the MDL. For example, in our DQO file, the threshold is 25 uS/cm. Currently the tabMWRacc() function is incorrectly using the MDL of 1 uS/cm. If I create a Lab Duplicate with result value 5, it still flags it as a miss.

[fawda123]

@ben-wetherill I think I've got these sorted out now (this commit https://github.com/massbays-tech/MassWateR/commit/8809a2385c3b9aa438a2444bc76ca106bbd65e05 and those referenced above). Can you please run your checks again to verify?

[ben-wetherill]

All examples are from new 6/9/22 results file.

From original list:

1. Field Blank/Lab Blank Quantitation Limit issue still exists - The MDL column in tabMWRacc(type = "individual") should show the Quantitation Limit if the result value is BDL. If the result value is not BDL, the MDL column should show the DQO MDL (or Lab Blank value). Looking at your code, I think you have it backwards. See example Field Blank, CND-036, 6/13/21, BDL, 0.02. The MDL column currently shows 0.01, but we need to know that the detection limit was not 0.01 in this case.
2. Good
3. Good
4. Good
5. Good
6. Good

New issues:

1. If the Field Blank or Lab Blank threshold is not equal to the DQO MDL, then tabMWRacc(type = "individual") should show the actual threshold in the MDL column. For example Sp Conductance. I suggest we also change the name of this column to "Threshold" to avoid confusion with the MDL in the DQO file.
2. Don't show results in scientific notation - See example tabMWRacc(type = "individual") Field Blank, CND-036, 6/15/21, 0.008.
3. In tabMWRacc(accck="Field/Lab Duplicates", type="individual"), the unit of measure for pH values shows up as "NA". Can we make that blank?
4. In tabMWRacc(type="individual") entries within parameter sections should be sorted by date. See example Lab Duplicate, TP, 8/14/21.
5. If values are less than the detection limit, they should still show as entered. Example Lab Duplicate, TP, 8/14/21, should show 0.0085 and 0.0095.
6. In the QC vignette issue list, we agreed to change the value for accchk-"Instrument Checks". But I also see that the header has been changed on tabMWRacc(accchk="Instrument Checks",type="individual"). I think we should keep that header as "Instrument Checks (post sampling)", because it is an important clarification for everybody.
7. In tabMWRacc(accck="Lab Spikes", type="individual") can we show more decimal places if necessary? There may be times when users enter more decimal places, such as if they are using mg/L when the data is closer to ug/L. See example Lab Spike, Ortho P, 2/17/21, 0.046 and 0.054. Might need up to 4 decimal places.

[fawda123]

@ben-wetherill thanks for the additional checks and issues. I think they're all addressed now, can you please verify when able? The new results file from 6/9/22 is also in the package.

[ben-wetherill]

In tabMWRacc(accck="Field Blanks", type="individual") there is still an issue with the value in the Threshold column. Example: E.coli, 9/13/21, ABT-162, 128, 50, threshold should show 10 because result value not BDL. Also, TP, 6/15/21, CND-036, 0.008, 0.001, threshold should show 0.01 because result value not BDL. Here is the logic for the threshold column:

DQO Thresh = DQO Field/Lab Blank if DQO Field/Lab Blank not BDL, else DQO MDL.
If result value is BDL, then...
        If Quantitation Limit not NA, then show Quantitation Limit
        Else show DQO Thresh.
Else show DQO Thresh.

And the logic for the Hit/Miss column:

If result value is BDL, then no MISS,
Else if result value < DQO Thresh, then no MISS,
Else, if result value = DQO Thresh AND DQO Field/Lab Blank uses <=, then no MISS,
Else MISS. 

From second list:

1. Good
2. Good
3. Good
4. Good
5. Good
6. Good
7. Good

New issues:

1. Field blank should be flagged as a miss if it is equal to MDL, because the DQO file says BDL. Example: tabMWRacc(accck="Field Blanks", type="individual") TP, 6/13/21, CND-009 0.01 should be a MISS.
2. Field blank should not be flagged as a miss if it is less than DQO Threshold. See logic above. Example: tabMWRacc(accck="Field Blanks", type="individual") TP, 6/15/21, CND-036 0.008 should not be a MISS.
3. In tabMWRacc(accck="Instrument Checks", type="individual"), the accuracy value for pH is not showing any decimals. It should show up to 2 decimals for pH, but for Sp Conductance, for example, decimals are not needed. Can you make the number of decimals float dependent on whether or not the difference requires them?
4. The same issue as #3 for the Diff/RPD column in Lab/Field Duplicates. We do not want decimals for the %RPD, but the Diff values do need to float. Can you make them float up to a maximum of 3 decimals? Example: Lab Duplicate pH values and TP values less than 0.05.

[fawda123]

@ben-wetherill all should be addressed now, hopefully I've got it sorted out! Please let me know if anything looks off.

[ben-wetherill]

The blanks threshold and hit/miss logic is working well.

From third list:

1. Good
2. Good
3. pH decimals look good. I misled you by saying that decimals are not needed for Specific Conductance. Sometimes they might be. In the new results file, see example Calibration Check, Spec Conductance, 8/12/21, 986.2. Accuracy should be 13.8, not 14.
4. Good

[ben-wetherill]

This looks good now.