Invalid results from the Retrieve Pixel Time Series Tool

[alexismc]

There is a discrepancy between the results provided by the Retrieve Pixel Time Series API command line tool, and the underlying data. The tool is not always producing valid results. An example is attached.

Example of invalid Retrieve Pixel Time Series results.docx

[sixy6e]

Not sure, I get the same results from the Retrieve Pixel Time Series Tool.

In [1]: import pandas

In [2]: df = pandas.read_csv('LS58_FC_WITH_PQA_WATER_118.25600_-22.25075_2000_01_01_2015_12_31.csv')

In [3]: df.set_index(pandas.to_datetime(df['ACQUISITION DATE']), inplace=True)

In [4]: df['2006-04']
Out[4]:
                    SATELLITE     ACQUISITION DATE  PHOTOSYNTHETIC_VEGETATION  \
ACQUISITION DATE
2006-04-17 02:00:45       LS5  2006-04-17 02:00:45                       6175
2006-04-26 01:54:43       LS5  2006-04-26 01:54:43                       5285
2006-04-26 01:55:07       LS5  2006-04-26 01:55:07                       5618

                     NON_PHOTOSYNTHETIC_VEGETATION  BARE_SOIL  UNMIXING_ERROR
ACQUISITION DATE
2006-04-17 02:00:45                           1656       2025            1801
2006-04-26 01:54:43                            891       3827            1827
2006-04-26 01:55:07                           1442       2906            1793

In [5]: df['2006-04-26']
Out[5]:
                    SATELLITE     ACQUISITION DATE  PHOTOSYNTHETIC_VEGETATION  \
ACQUISITION DATE
2006-04-26 01:54:43       LS5  2006-04-26 01:54:43                       5285
2006-04-26 01:55:07       LS5  2006-04-26 01:55:07                       5618

                     NON_PHOTOSYNTHETIC_VEGETATION  BARE_SOIL  UNMIXING_ERROR
ACQUISITION DATE
2006-04-26 01:54:43                            891       3827            1827
2006-04-26 01:55:07                           1442       2906            1793