Sandor Kertesz

### What happened? File `c_85.bufr` contains a single BUFR message with 2 `airTemperature` values. The structure is as follows: When we try to extract the `airTemperature` with the following code:...

### Is your feature request related to a problem? Please describe. Currently, unpacking a message internally goes like this: ```python msg["unpack"] = 1 ``` and message object is required to...

Add missing value handling options

6

Currently `read_bufr` does not offer control over missing values during the extraction and we have to filter the resulting Pandas dataframe to remove them. Option 1 --------- Add option `...

### Is your feature request related to a problem? Please describe. When we create a PR in `pdbufr` and push new code into it two sets of CI tests are...

### What happened? Using e.g. SYNOP data: ``` df = pdbufr.read_bufr(columns=["latitude", "longitude", "airTemperatureAt2M"], filters={"WMO_station_id": [30846, 89514] ``` does not apply the filter but extracts data from all the messages. To...

### Is your feature request related to a problem? Please describe. Related to #51 The data in question has the following structure: And we would like to perform: ``` df...

Support rank notation in ecCodes keys

1

ecCodes supports the **#n#key** notation to get the **n**-th value for a given **key** in a BUFR message. It would be great if pdbufr allowed for this notation. E.g this...

Improve filter speed

2

The performance of the bufr filter should be improved. It is currently **4-5 times slower** than the BUFR filter in Metview Python (it is based on a C++ wrapper around...

This would make the comparison to Metview's bufr filter possible.